SUSCEPTIBILITY OFSERVICE EXPOSED CREEP RESISTANT …

SUSCEPTIBILITY OF SERVICE EXPOSEDCREEP RESISTANT MATERIALS

TO REHEAT CRACKINGDURING REPAIR WELDING

Submitted in partial fulfillment of the requirements for the degreeMasters in Metallurgical Engineering in the Faculty of

Engineering, the Build Environment andInformation Technology

©© UUnniivveerrssiittyy ooff PPrreettoorriiaa

TABLE OF CONTENTS

CHAPTER1 1

1.1 Introduction 11.2 Cr-Moand Cr-Mo-VSteels 21.3 References 5

CHAPTER2: LITERATURESURVEY 7

2.1 The HeatAffectedZone (HAl): Transformationand Behaviour 72.2 ResidualStresses in a Weldment 92.3 MicrostructuralEvolutionin the HAl upon PWHT 112.4 ReheatCracking 142.5 MicrostructuralEffects 172.6 CompositionalEffects (AlloyChemistry) 182.6.1 Effect of Chromiumand Molybdenum 202.6.2 The effect of Vanadium 252.6.3 The effect of Carbon 272.6.4 The effect of Nickel 272.6.5 The effect of Manganeseand Silicon 272.6.6 The effect of Titanium 282.6.7 The effect of Trace elements 282.7 Ways to mitigate reheatcracking 292.8 References 30

CHAPTER3: EXPERIMENTALPROCEDURE 34

3.1 ChemicalAnalysis 383.2 CGHAZSimulation 383.3 SampleManufacturing 393.4 Artificial Ageing of P91Material 393.5 Instrumentsand MeasurementsduringTesting 393.6 ScanningElectronMicroscopy 393.7 Metallography 403.8 HardnessSurvey 413.9 References 41

CHAPTER4: %Cr-%Mo-Y.V 42

4.1 Introduction 424.2 ChemicalAnalysis 424.3 Spiral NotchTest Results 434.4 SEM Fractography 454.5 Metallography 484.6 HardnessSurvey 544.7 Discussion 544.8 Conclusion 554.9 References 56

CHAPTER 5: 2Y.•Cr-1 Mo

5.1 Introduction5.2 ChemicalAnalysis5.3 SpiralNotchTest Results5.4 SEMFractography5.5 Metallography5.6 HardnessSurvey5.7 Discussion5.8 Conclusion5.9 References

CHAPTER 6: X20CrMoV12-1


CHAPTER 7: P91


57

575758606369697171

72

727273757782828484

85

858586889096969898

High temperature steam generating boilers constructed from creep

resistant alloys are designed for specific operating parameters for the

estimated safe operating life. Design life calculations of a boiler are

based on material data available at that time and also on previous

experience. During the life of the boiler new material data becomes

available through long-term research projects, in some cases showing

weak characteristics and behaviour that can compromise the original

design life. That plant operator needs to launch a remedial maintenance

and inspection programme or review the operating parameters to ensure

safe and reliable remnant life for the affectep components. A significant

proportion of Eskom's boiler fleet accumulated more than 100 000

operating hours thus requiring close evaluation of component health in

view of life extension beyond the original design life estimates.

German Spec British Spec Also known asDIN17715 BS3604 Grade

HFS:15M03 - %Mo

13CrM044 620 1Cr-%Mo10CrM0910 622 2~Cr-1Mo14MoV63 660 %Cr-%Mo-~V

X20CrMoV12 1 762 12CrorX20X10CrMoVNb91 Steel 91 or P91

Critical steam containing components manufactured from a combination of the

above alloys accumulated sufficient operating hours for creep damage to take

appreciable effect. In theory it would be possible to safely operate material

which is creep aged to the upper limit of the secondary creep phase, just

before the rapid creep deformation associated with tertiary creep commences.

Research work included specifically the susceptibility of four of the above

mentioned materials (%Cr-%Mo-~V, 2~Cr-1Mo, X20, P91) with regards to

reheat cracking, as these types of materials, especially the older ferritic

grades (%Cr-%Mo-~V and 2~Cr-1Mo) are prone to reheat cracking. The

tests involved the material undergoing a simulated welding thermal cycle and

then being subjected to constant load tests. This was done for new as well as

service exposed material.

The results indicate that for service exposed materials where creep damage

(voids in the microstructure) are below a certain limit the material show no

susceptibility to reheat cracking for the testing conditions employed. For the

new material it was confirmed that mainly %Cr-%Mo-~V showed susceptibility

towards reheat cracking.

Stoom ketels, in die elektrisiteits generasie industrie, word vervaardig

van kruip-bestaande stale en is ontwerp vir spesifieke operasionele

parameters vir die beplande, veilige operasionele leeftyd van die ketel.

Ontwerpslewe berekeninge van so 'n ketel word gebasseer op

beskikbare materiaal inligting op daardie tydstip en ook op vorige

ondervinding van die ontwerpspan. Tydens die leeftyd van 'n ketel word

nuwe materiaaldata dikwels bekend gestel. Hierdie resultate is as

gevolg van lang termyn navorsings projekte. In sommige gevalle word

swak materiaal eienskappe uitgewys wat 'n sterk invloed het op die

ontwerp leeftyd van die ketel. Aanlegbestuur en operateurs het dus

nodig om 'n voorkomende inspeksie en onderhoudsprogram in te stel of

'n volledige studie van aanlegparcimeters moet gedoen word en dalk

aangepas word, om te verseker dat geaffekteerde komponente veilig

gebruik kan word vir die res van die leeftyd. 'n Relatiewe groot

hoeveelheid van Eskom se stoomketels het al meer as 100 000

operasionele ure geakumuleer. 'n Filosofie is in plek gebring wat baie

deeglike inspeksie van komponente vereis, met die oog op verlenging

van die operasionele leeftyd, verby die oorspronklike ontwerp leeftyd.

Kruip bestande legerings soos gebruik op Eskom aanlegte wordaangedui in Tabel 1:

Duitse Britse Spesifikasie Ook bekend asSpesifikasie BS3604GraadDIN17715 HFS:15M03 - ~Mo

13CrM04 4 620 1Cr-~Mo10CrM0910 622 2~Cr-1Mo14MoV63 660 ~Cr-~Mo-~V

- X20CrMoV12 1 762 12CrorX20X10CrMoVNb91 Staal 91 of P91

Kritieke stoomdraende komponente wat vervaardig is van 'n kombinasie van

die bogenoemde legerings, het al genoeg operasionele ure geakumuleer vir

kruip skade om te manifesteer in die mikrostruktuur. In teorie sal dit moontlik

wees om die komponente veilig te gebruik tot in die boonste limiet van die

sekondere kruip fase. Dit is tot net voor hoe kruip vervorming wat geassosieer

word met tersiere kruip begin.

Navorsingswerk tydens hierdie studie het gefokus op die vatbaarheid van vier

van die bogenoemde legerings (Y2Cr-Y2Mo-~V,2~Cr-1Mo, X20 en P91) vir

herverhittings krake. Hierdie tipe materiale, veral die ouer ferritiese legerings

(Y2Cr-Y2Mo-~V en 2~Cr01Mo), is vatbaar vir herverhittingskrake tydens

sweiswerk. Die toetse het behels dat die materiaal 'n gesimuleerde termiese

sweis siklus ondergaan en dan onderwerp word aan konstante belastingI

toetse. Dit was uitgevoer op nuwe sowel as diensverouderde materiaal.

Die resultate het aangedui dat vir diensverouderde materiaal waar kruip

skade (holtes in die mikrostruktuur) onder 'n sekere hoeveelheid is, die

materiaal nie vatbaar is vir herverhittingskrake nie, vir die toets kondisies wat

gebruik is. Vir nuwe materiaal was dit bevestig dat Y2Cr-Y2Mo-~Vdie meeste

vatbaar is vir herverhittingskrake.

I would like to thank Professor G.T. van Rooyen for his invaluable guidance

and input for the duration of this project.

Professor P.C. Pistorius, Mr J. Borman and Mr C. Smal from the Department

of Metallurgical Engineering for always being available and offering

assistance.

,Mr P Doubell, a colleague from Eskom TSI for his help and guidance.

Mr G von dem Bongart, for his assistance with the various microstructures

obtained during the experimental work.

Mr C Coetzee, for his assistance with the scanning electron microscope

(SEM).

Eskom for provision of material and financial aid, as well as the opportunity to

conduct this task.

The main energy supplier, world wide today, for electricity is still fossil fuel

power stations. A huge quantity of the construction in these power stations is

done by means of fusion welding processes. The wide range of materials

used to build these huge constructions all have to meet rigid specifications of

high strength, good toughness and often resistance against fatigue, corrosion

and creep deformation (high temperature areas like the boilers and main

steam pipelines).

However the most important factor in the choice of these materials is

essentially that they posses good we/dability. The question arises then: What

is meant by 'good we/dability'? This can be defined as the ability of a material

or combination of materials to be welded under fabrication conditions into a

specific suitably designed structure and to perform satisfactorily in intended

service1• As such the concept of good weldability has to be a function of a

number of interacting variables, which include 2:1. Type of welding process

2. Composition and mechanical properties of weld metal

and base metal

3. Welding sequence

4. Energy input/

5. Joint design and size

6. Welding consumables

7. Structural constraint

8. Environment

A very simple definition of weldability is more often used. Will the material be

free from cracking during fabrication and subsequent use? A clear distinction

has been drawn between cracking occurring during the service life of the

material and cracking occurring very shortly after welding or during welding.3

One such type of the latter mentioned is reheat cracking, also called stress

relief cracking.

It has been demonstrated that reheat cracks form by a low ductility creep

cavitation mechanism at the prior austenite grain boundaries during post weld

heat treatment (PWHTt Thus reheat cracking usually occurs in the coarse

grained heat affected zone (CGHAZ) and by means of intergranular fracture

along the prior austenite grain boundaries. A troubling aspect encountered

during reheat cracking is heat-to-heat variations, with respect to cracking

susceptibility of the material. This indicates dependence on residual elements

rather than on the bulk chemistry. The bulk chemistry of the alloys may

therefore not necessarily be a reliable predictor of susceptibility to cracking.5

At present the phenomenon and mechanism responsible for reheat cracking

has been well researched and documented. Nevertheless, reheat cracking is

still frequently a problem during welding, necessitating expensive repairs and

reheat treatment. Some empirical guidelines on how to prevent reheat

cracking have been established over the years. A much better understanding

of the complex interaction of welding stresses, weld defects, geometrical

constraint and material properties are required. This is of particular concern in

the repair welding of creep damaged components, which may be at a greater

risk of cracking than new material.

The properties sought after in a steel grade for high temperature applications

are the following:

• High hot tensile strength

• Low creep rate

• High stress rupture ductility

• Corrosion and/or oxidation resistance

Cr-Mo and Cr-Mo-V steels are designated for use at elevated temperatures

because of their resistance to creep. They are used as construction material

for the construction of steam power plants, especially boilers, high pressure

pipes and tubes, flanges, forgings, turbine rotors, water boilers, collectors and

other highly stressed structural components.6

These steels typically contain 0.5 - 9.0% chromium and 0.5 - 2.0%

molybdenum, with a carbon content usually less than 0.2%. Vanadium is

sometimes added in the range of 0.35 - 0.7% and provides added creep

strength due to formation of fine vanadium carbide precipitates. Small

additions of nickel also improve the toughness and ductility. Chromium and,

molybdenum, when used as alloying elements in steel increases the

hardenability and strength at elevated temperatures. Chromium further also

improves resistance to oxidation and scaling. The most popular grades of Cr-

Mo and Cr-Mo-V steels used today are listed in Table 1.1.

German Spec British Spec Also known asDIN17715 BS3604 Grade

HFS:15M03 - l'2Mo

13CrM044 620 1Cr-l'2Mo10CrM0910 622 2'!4Cr-1Mo14MoV63 660 l'2Cr-l'2Mo-'!4V

X20CrMoV121 762 12CrorX20X10CrMoVNb91 Steel 91 or P91

Cr-Mo steels are normally used in a heat-treated condition. The heat

treatment is either quenching and tempering or normalising and' tempering.

Sometimes 'annealing or isothermal annealing is used to produce a soft,. -relatively more stable ferritic microstructure. Two of the factors that make Cr-

Mo steels so attractive for high temperature applications are transformation

hardening, but more important is the fact that low alloy Cr-Mo steels have

resistance to softening during post weld heat treatment (PWHT) and creep

service. The maximum resistance depends on the degree of precipitate

dispersion. Chromium can form several types of precipitates. Chromium is

soluble in cementite-M3C and it tends to spheroidize these carbides. Three

additional carbides can also be formed - Cr23CS(cubic), Cr7C3(hexagonal)

and Cr3C2(orthorhombic), each of which can dissolve considerable amounts

of iron. The extent of solid solubility of iron in Cr7C3is significant and thus it

prevents the third carbide, Cr3C2,from occurring in equilibrium with ferrite.

Molybdenum can form hexagonal M2C carbides, which are not in

thermodynamic equilibrium with ferrite. Therefore, during long exposure times

at elevated temperatures, MsCtype carbides evolve 7.

I

Vanadium carbide, V4C3 (cubic), is the most important carbide in Cr-Mo-V

steels. The improvement of the cre~p strength of Cr-Mo-V steels over Cr-Mo

steels is due to the presence of very fine vanadium carbides, in the matrix,

that are more stable than either chromium and molybdenum carbides 8.

As mentioned before these steels are widely used in the manufacture of

components that operate in the creep regime, such as boiler waterwall tubes,

headers and main steam pipes, in power stations. The superheated steam

(565DC) is transported from the headers to the turbine by the main steam

pipeline. These pipes carry large volumes of steam and like the headers are

not subjected to the aggressive environment of the boiler itself.

The dimensions of the different pipes are c91lculatedon the basis of a service

life in excess of 100 000 h (11% years). The useful life of these pipes is more

often actually determined by the development of creep damage, especially

near or in the weldments. The creep damage is often characterised by the

development of a number of microcavities in the microstructure of the steel.

These microcavities may eventually coalesce into creep cracks. Inspection of

replicas of the microstructure is an acceptable method to assess the remnant

creep life of boiler pipe structures.

Creep damage is in no way uniform throughout the pipes. Some areas,

especially in or around weldments may be severely damaged while the

surrounding base metal may contain very little damage. Therefore it has

become practice to replace only the damaged sections. To date the

acceptable creep damage in the parent metal has been set at 50 voids/mm2.

Therefore if 50 voids/mm2 or more are detected, the section is no longer fit for

service and has to be replaced. A new section will then be prepared and

welded into position. This implies that new material and service exposed

material will be welded together.

This raises the following questions:

1. To what extent will the prior creep damage be affected by the welding

thermal cycle?,

2. What effect will this prior creep damage in the HAZ have on the

integrity on the weldment?

3. What effect will the subsequent thermal exposure have on the integrity

of the weldment?

4. Will the service-exposed material be more susceptible to reheat

cracking during post weld heat treatment (PWHT), after welding?

1. 'Welding Metallurgy", Welding Handbook, ih Edition. Vol. 1, Chapter 4, American

Welding ~ociety, 1982.

2. Luther, G.G., Jackson, C.E. and Hartbouer, C.E., "A Review of Weldability Test of Carbon

and Low Alloy Steels". Welding Journal, Research Supplement, October 1949, pp. 376s-

396s.

3. Coe, F.R, "The avoidance of Hydrogen Cracking in Welding", The Welding Institute,

1973.

4. BonisZewski, T., ISI Conference on "Heat Treatment Aspects of Metaljoining Processes",

London, 1971, pp. 29-41.

5. MCPl'!erson,R, "Compositional effects on Reheat Cracking of Low Alloy Ferritic Steels",

Metals Forum, Vo1.2(3),1980, pp. 175-186.

6. Wegmuller, C.R, "Chrome-Moly Steel Builds Big Vessels", Welding Design and

Fabrication, July 1983, pp. 47-52.

7. Lundin, C.D. and Khan, KK, "Fundamental Studies of the Metallurgical Causes and

Mitigation of Reheat Cracking in 1XCr-Y:zMoand 2XCr-1Mo Steels", WRC Bulletin 409,

February 1996.

8. Parker, J.D. and Parsons, A.W.J., "High Temperature Deformation and Fracture

Processes in 2~Cr1 Mo-Y:zCrY:zMo~VWeldments", Int. J. Pres. & Piping 63, pp. 45-54,

1995

2.1 THE HEAT AFFECTED ZONE (HAZ): TRANSFORMATION ANDBEHAVIOR

When materials are joined by fusion welding processes, parts of the material

have to be heated up to its melting point (Tm). This severe thermal cycle

causes transformation of the original microstructure and changes in the

properties of the material, in a region close to the weld. This volume of metal,

or zone, is usually referred to as the heat affected zone (HAZ).

The metallurgical transformations that occur during welding affect the final

microstructure and therefore may influence many of the problems that can

develop during and after welding. The coarse grained heat affected zone

(CGHAZ) is the location of maximum susceptibility for reheat cracking. It is

also the primary region for reduced toughness.

Results from Ito and Nakanishi1 indicates that in 1Cr-%Mo alloys a HAZ

consisting of martensite or lower bainite was more susceptible to reheat

cracking than upper bainite. Thus, the HAZ transformation characteristics play

a vital role in the weldability of a material. The tra"nsformationbehaviour may

in turn provide the key to reducing or eliminating weld HAZ problems.

/Easterling2 has compared regions in the microstructure in a weld with the iron-

carbon (Fe-C) equilibrium phase diagram (Figure 2.1). Accordingly the HAZ

can be divided into sub-zones. Each sub-zone has its own distinct

microstructure, as well as mechanical properties. A few problems arise with

this simple type of representation. The weld thermal cycle differs significantly

from conditions used to obtain the phase diagram, especially heating

(1600°C/s) and cooling rates (260°C/s). Also, during welding, complete

homogenisation never exists. Non-equilibrium constituents such as martensite

and bainite are also not taken into consideration.

:~~-; , ~lOl>O••~.C _.~.i \ '.Old: • ,',<:",., ". '" "",,. ,"" •••tjl ".

: I \I \ '.Cf'tt.laJ~;l.~;tQo"'.

\ \I '\hH'4H., 'f.~lt~M.i:

ion.e

Figure 2.1: A schematic diagram of the various sub-zones of the heat affected zoneapproximately corresponding to the alloy Co (0.15wt % C) 2.

The same objections to the use of the equilibrium phase diagram, to predict

weld HAZ transformations, also extend to the use of standard continuous

cooling transformation (CCT) diagrams. The starting base when developing

these diagrams is homogeneous austenite. In welding, there is very little to no

homogeneity, due to the inability of alloying elements to diffuse uniformly

throughout the austenitic phase and the incomplete solution of carbides,

nitrides and other constituents. This is due to the rapid heating and cooling

rates of the weld thermal cycle and the con~omitant short austenitizing times.

In order to accurately predict the on-cooling transformation temperatures and

microstructures, weld HAZ CCT diagrams must be derived using the heating

and cooling conditions relevant for welding. Very few of these CCT diagrams

have been developed under welding conditions. Conventional CCT's are. -however available for most of the commercial grades of Cr-Mo and Cr-Mo-V

steels.

When these conventional CCT's are studied it becomes apparent that the

resulting microstructure under various welding conditions may be complex.

Depending on the degree of homogenisation and the cooling rate (related to

the heat input and preheat for a given process and material thickness), the

on-cooling microstructures in the weld HAZ might consist of martensite, mixed

martensite and bainite or bainite coupled with retained austenite.

The following have been found when 1~Cr-%Mo and 2~Cr-1Mo was studied.

In regions containing homogeneous or nearly homogeneous austenite

(regions heated above 1100°C) the ferrite reaction is suppressed and only a

bainitic reaction occurs. The reaction start temperatures are in the vicinity of

540°C, depending on the peak temperature reached and the grain size of the

austenite. The regions heated between 955-1100°C contain undissolved

carbides'. These carbides can act as nucleating sites for the formation of pro-

eutectoid ferrite in addition to the bainite reaction products. In the regions of

the HAZ heated in the range 790-955°C, austenitization is limited to only

those areas immediately surrounding the grain boundaries. This continuous

network of austenite may transform to martensite that can have poor impact

properties. These trends are found in general in the Cr-Mo and Cr-Mo-V

steels and again shows the complexity of the resulting HAZ microstructure

after welding these materials.

Residual stresses develop during cooling in the weld HAZ and fusion zone

due to restrained shrinkage and transformation volume changes as a result of/

austenite decomposition. On cooling, those areas of the base metal that

experienced thermal expansion due to heating must contract or flow

plastically. The bulk of the base metal that has experienced no significant

heating (and therefore no decrease in strength) prevents or restrains the

contraction of the cooling material. Above approximately 650°C, the weld

fusion zone and those areas immediately adjacent to the weld accommodate

the thermal contraction by plastic deformation without developing any

significant stress. This is the due to the fact that the yield strength is low

above this temperature. Cooling below 650°C results in significant increases

in yield strength with decreasing temperature. Plastic deformation only occurs

~-~--"-+-,....'. :-1 \I ., .\

Temp /only

...~ ... t ..•.•i -I \ )(. , "Tronslormotl(lt'\

. only

Figure 2.2: Schematic illustration of transverse residual stresses at the weld centre linecaused by interaction of shrinkage, quenching and transformation strains3

•

when stresses due to thermal contraction exceed the yield stress. Therefore,

cooling to the preheat temperature results in increasing residual tensile stress

concomitant with increased yield strength in the fusion zone and HAZ. The

resultant residual tensile stresses that occur in the HAZ and fusion zone are in

force and moment equilibrium with comprehensive stresses in the bulk of thebase material.

Transformational stresses are a result of the volumetric expansion that occurs

during decomposition of austenite. The material being transformed attempts

to expand but expansion is hindered by the cooler material that is not

undergoing transformation. The material being transformed therefore

experiences a compressive stress and the cooler material a tensile stress. If

the transformation temperature is high, the effects of subsequent bulk

shrinkage will override the transformation stresses. However, if the

transformation temperature is low, the transformation stresses will lower the

overall tensile stress in the HAZ and fusion zones.

Superposition of the components of the residual stress developed during

welding lead to an extremely complex final residual stress state. The CGHAZ,

that portion of the HAZ adjacent to the fusion zone, is subjected to a

significant tensile stress both in the direction of and perpendicular to the weld.

These biaxial residual stress states as well as the unfavourable metallurgical

characteristics of the CGHAZ exacerbate the susceptibility of this zone to

reheat cracking.

,A schematic diagram of transverse residual stresses is shown in Figure 2.2.

A major function of PWHT is to restore ductility in the HAZ and weld metal in

Cr-Mo and Cr-Mo-V weldments. In addition, the PWHT also reduces the

residual stresses in the weldment by a creep relaxation process.

Detailed recommended practices for welding Cr-Mo and Cr-Mo-V steels are

given in ANSI/AWS 010.8-864. In this code the various PWHT temperatures

and holding times for different grades of Cr-Mo steels are specified.

During austenitization in a weld thermal cyc;le,all or part of the carbides are

taken up into solution, depending on the peak temperature experienced, the

material thickness and energy input. During the subsequent cooling the matrix

transforms to bainite/martensite/ferrite. The end product still remains

supersaturated with respect to carbon as well as alloying elements that

subsequently precipitate as carbides during tempering. The various types of

carbides present in Cr-Mo and Cr-Mo-V steels are MC, M2C, M3C, M7C3 and

MeC. The carbide type, size, distribution and morphology will depend on the

chemical composition, microconstituents present at the temperature and time

of tempering.

Baker and Nutting5 have shown that the types of carbides present in 2%Cr-

1Mo base metal are dependant on, starting microstructure, heat treatment

(tempering) and time at the tempering temperature. After normalising, the

microstructure was found to consist of a mixture of ferrite and bainite whereas

after quenching the microstructure was mainly bainitic. They determined that

the carbide evolution in the bainitic regions of both quenched and normalised

material was similar as shown below:

E-carbide t't cementite+ --.. cemen Ie --.. + --..

cementite M02C

________ • M7C3

IM23Ca--.. MaC

In the ferritic regions the M2Ctype c.arbidesdo not go through all the transition

carbides and transform directly to MaC carbides. The M2C carbides were

found to be more stable in the ferrite than in the initial microstructures

consisting of martensite or bainite. In the martensitic and bainitic

microstructures M23Cagrew at the expense of M02C, degrading the creep

properties with increasing tempering temperature or time at tempering

temperature. This has led some researches to recommend using materials in

the normalised and tempered condition rather than in the quenched and

tempered condition. The presence of pro-eutectoid ferrite is however

generally regarded as unfavourable, due to its poor toughness and low

strength. Current alloy development reflects the need to avoid the formation to

primary ferrite in the microstructure duriryg cooling by making increased

hardenability one of the principal criteria in alloy design.

Modification of Cr-Mo alloys by the addition of vanadium results in the

precipitation of finely dispersed, slow growing V4C3 upon tempering or post

weld heat treatment. The precipitation occurs directly inside the matrix and

unlike the precipitation of M02C, is not dependent on the formation of any

other carbides. In alloys containing more than 0.1 % V, the M02Ccarbides can

have a metal atom ratio of up to 0.3 of Vanadium present. Vanadium in M02C

results in the stabilisation of M02C.The combined effects of precipitation and

stabilisation result in vanadium being one of the most potent elements in

promoting creep resistance.

One is inclined to believe that the carbide evolutionary sequence in the

CGHAZ may be similar to that in the base metal. The situation is however

complicated by the fact that some of the more stable carbides such as TiC,

NbC and V4C3, although finer than the chromium, molybdenum or iron

containing carbides, may not dissolve. Lundin et a/.6 have found that in the

2~Cr and 3Cr alloys that contain modifying elements such as vanadium,

niobium and titanium, the carbides do not completely dissolve during a coarse

grain HAZ simulation thermal cycle. However, in standard alloys, all the,

carbides dissolve upon CGHAZ simulation. Thus the subsequent precipitation

of carbides during PWHT will be affected. The investigation further revealed

that in steels modified with vanadium, titanium and boron, the M02C type

carbides persist for longer times when compared to other alloys at a PWHT

temperature of 675°C. The carbide evolution sequence in the CGHAZ of

1CrMoV steel is shown in Figure 2.3. It can be seen from Figure 2.2 that the

carbides present in the CGHAZ on tempering will depend on the PWHT

temperature.

MATERIA["WW_W" HAZ COAR~E:·GRAlNw.w-wwm-------sn=fessGSi7-CrMoV SIMULATION RELIEVING

T=1300°C. t ~:125 AT T~C:2

MX

MzeMnC,

M,C,

AS WELDED STRUCTURE TEMPERED HAZ • STRUC1'URE

FERRITE

II+-~:

725 'C

IMX.M:C.M.C,.MuC.!700·C

(M,C,MX.M:C.M1,C, I550·C

IM,C.MX,M:C550·C

1 M,C.MX

Several reasons may exist to reheat the weld metal after welding; to relieve

residual stresses in the weld and to obtain the best possible microstructure in

the HAZ and the weld metal. The principal problem to be studied here is

reheating of the weld in the range 500-650°C in order to obtain stress

relieving, and the type of cracking that may result from this treatment.

Cracking is manifested by low rupture ductility and intergranular fracture along

the prior austenite grain boundaries. An example of reheat cracking in a multi-

run butt weld in Cr-Mo-V steel is shown in Figure 2.4.

Figure 2.4: Reheat crack in Cr-Mo-V butt weld. After Branah, G.D., High temperature

mechanical properties as design parameters.

During the post weld heat treatment (PWHT), the residual stresses are

r.elaxed by localised creep deformation. Any part of the weld, like notches and

sharp transitions, which are in a state of high tensile residual stress, will act as

strain raisers, when the stresses are relaxing at high temperatures. Cracking

and fissuring can then easily nucleate in a susceptible material.

As stated above the reheat cracks form on the prior austenite grain

boundaries, in that part of the HAZ that has experienced temperatures well

above -1100°C, and subsequently has undergone plastic deformation.

Sometimes the cracks are well developed (optically visible and extending

along several grain boundaries), and other cases the damage appears only as

a row of small voids on the grain boundaries.

It is now believed that several conditions must be present for reheat cracking:

1. The microstructure must be susceptible - a coarse prior austenite size

like the CGHAZ of welds.

2. Residual stresses must be present.

3. Stress concentrators, like notches or discontinuities - weld

configuration, slag inclusions, lack of fusion and cracks.

It has been shown that if a material containing a CGHAZ, undergoes a post

weld heat treatment, without any plastic strain occurring during this heat

treatment, no loss in elevated temperature bend or tensile ductility is shown.

Simulated HAZ microstructures do show an increase in impact transition

temperature, even without plastic strain. Reheat and stress rupture cracks

form because the relaxation strains exceed the creep ductility of the CGHAZ

at the applicable temperatures.

A brief overview of some of the theories regarding reheat cracking will now

follow:

2.4.1 When the temperature exceeds 1200°C, carbides like chromium,

molybdenum and vanadium are taken into solution in. the HAZ.

Coinciding with this high temperature interval, grain growth occurs in

tne HAZ. The rapid cooling from these high temperatures and the low

transformation temperatures of these steels suppresses the

reprecipitation of the carbides. The final microstructure, after welding,

is usually martensitic or even lower bainite. During the PWHT and/or

the service life the carbide forming elements precipitate out of the

supersaturated solid solution as carbides. The manner in which the

precipitation occurs is similar to that encountered during normal

tempering. The precipitates form in the grain interior and although

initially small in size (a few Angstroms in diameter), cause 'stiffening' of

the grain interior. Secondary hardening occurs depending on the

temperature. At the grain boundaries there are not precipitates of a

sufficient size to prevent grain boundary sliding. It has also been

observed that some areas next to the grain boundaries may contain no

precipitates (denuded zone)8. Plastic deformation imposed on a

microstructure in this condition will be restricted by the strong (hard)

grain interiors and subsequently most of the deformation will take place

along the weaker grain boundaries or denuded zones, causing grain

boundary sliding. During the stress relieving, the overall strains may be

small, but with a large grain size (like the CGHAZ) there is less grain

boundary area, and local high shear and tensile strains develop at the

grain boundaries. The resulting significant deformations lead to

formation of voids at discontinuities on the grain boundary interfaces

and these cavities when linked up, form the final grain boundary

cracks.

2.4.2 Trace elements or impurities playa very important role in reheat

cracking. It has been shown that high purity heats of the same bulk

chemistry do not show a ductility loss in simulated CGHAZ

microstructures. However various interactions exist among the alloying

elements and the trace elements can affect ductility. This is most likely/

due to the segregation of trace impurities to the grain boundaries at

high temperatures and this can cause embritllement. Several

researchers have found that segregation of residual elements (sulphur,

phosphorus, tin and antimony) cause severe problems by embritllingtile grain boundaries.9,10,11 In work done by Hippsley et a/.12 it was

noted that the segregation of impurities at the grain boundaries can be

due to equilibrium segregation from the grain matrix or solute rejection

from grain boundary carbides. It has been stated that although the

segregation of trace elements at elevated temperature plays a major

role in the loss of ductility, the closest to the true mechanism is most

likely a combination of the two ideas - precipitation strengthening andsegregation.9,1o

The necessary factors for reheat cracking was summarised by Ito andNakanishi: 1

1. The material must have undergone a thermal cycle that results in

solution of alloying elements and retains the elements in solid solution

after cooling.

2. Grain growth must have occurred as a result of thermal cycling.

3. Heat treatment within the temperature range 450-700°C, resulting in

significant precipitation strengthening.I

4. Grain strength and internal stress must exceed the strength of the grain

boundaries.

5. A stress riser must be present to initiate cracking.

sa c•.••••••""9

Gro_ ~f'Y .tid"-9

5••.••"'91 •• oS 9#""'" ~ <. •.••••••...••.1 ••••.•n-.:::•.... "'9... of 0"- _

Researchers have shown that a martensitic or lower bainitic microstructure is

more susceptible to reheat cracking than upper bainite. However, they state

that this difference is related to the precipitation processes, rather than the

optically resolved microstructure. The martensitic and lower bainitic structures

are supersaturated with alloying elements and carbon. During the PWHT this

leads to severe secondary hardening (precipitation of carbides).1,13

The composition of these high temperature application steels plays a vital role

with regard to the susceptibility to reheat cracking. For the high temperature

application steels these alloying elements are usually chromium, molybdenum

and vanadium. They increase the tensile and creep strength due to carbides

or carbonitride precipitates in the ferrite matrix. Unfortunately according to

many researchers these elements also greatly enhance the susceptibility to,

reheat cracking. To restrict these elements to very low alloy additions is not

practical because their presence is important for the hardenability, strength

and creep resistance of these steels.

The following elements were found to be detrimental to cracking: carbon,

vanadium, molybdenum (individually and in concert with vanadium), niobium,

aluminium, and copper. With regards to residual elements: phosphorus,

sulphur, tin, antimony and arsenic. Chromium, boron and titanium's effects

are not as clearly defined. Nickel was found to have no effect on cracking. To

simplify, the elements that form or promote the M2C or M4C3 type carbides, or

enhance grain boundary embrittling are deleterious.

Over the years many researchers have atteJllPted to quantify the effect of the

alloying elements in different alloys on susceptibility to reheat cracking.

Nakamura el al.14 made such an attempt to determine the effect of the alloying

elements in Cr-Mo steels. Their work resulted in the L\G· parameter

relationship.

(6-1 )

Ito and Nakanishi1 extended Nakamura's work, by including additional alloying

elements like nickel, copper, niobium and titanium. Their work resulted in

cracking parameter PSR:

PSR= Cr + Cu + 2Mo + 10V + 7Nb + 5Ti -2 (6-2)

If PSR> 0, the material is considered susceptible to reheat cracking. However,

the PSR parameter is only applicable to alloys containing less than 1.5%Cr,

2%Mo, 1%Cu and 0.15%V, Ti and Nb. They also claimed that a chromium

content larger than 2% eliminated cracking. The amounts of Ni and Cu added

seemed to have no effect on susceptibility and do not appear in the

parameter.

,However, many researchers have found a poor correlation between these

parameters and actual susceptibility to reheat cracking, of different alloys.15,16

It has also been found by researchers that 214Cr-1Mo alloys are susceptible

to reheat cracking although Ito and Nankanishi1 claimed that a chromium

content of 2% or greater eliminated cracking.

McMahon et al.16 suggested another parameter:

CERL + Cr = Cr + 0.2Cu + 0.44S + P + 1.8As + 1.9Sn + 2.7Sb (6-3)

It is quite clear that trace elements and not carbide formers are the

detrimental elements in this parameter; trace elements causing embrittling of

the grain boundaries. The greater the CERL + Cr value, the higher the

susceptibility to reheat cracking.

Boniszewski17 suggested the use of a metal composition factor (MCF) :

MCF = Si + 2Cu + 2P + 10As + 15Sn + 20Sb (6-4)

The MCF also focuses on the grain boundary embrittling elements. An

increase in the MCF correlates with a decrease in rupture ductility as

measured by elongation in hot tensile tests.

~G1 = Cr + 3.3Mo + 8.1V + 1OC - 2

If ~G1 > 2, the material is considered susceptible to reheat cracking.

Considering the effects for trace elements alone, the R-value was developed

for O.5CrMoV steels:

R = P + 2.43As + 3.57Sn + 8.16Sb (6-6)

Susceptibility to reheat cracking increases with an increase in R-value.

Bruscato 18 devised an embrittling factor relating to weight percent of impurity

elements (in ppm.):

x = 10 P + 5 Sb + 4 Sn + As100

Susceptibility increases with X values.

In the subsequent section the effect of some alloying elements individually

and in combination with other elements are reviewed.

As seen by its appearance in the i1G and PSR parameters, chromium

increases the susceptibility of a material to reheat cracking. However,

according to Ito and Nakamura a chromium content of larger than 2% should

eliminate reheat cracking. The published literature does not agree with this as

reheat cracking has been found in material with up to 3% chromium. This

could be due to the effect of other elements such as molybdenum, vanadium,

titanium and niobium in the steels. Some results show that chromium between

o and 2% decreases the high temperature ductility and values above 2%

increases it markedly.

Molybdenum also appears in the i1G and PSRparameters and it is clear from

these Rarameters that it has a greater effect on susceptibility to reheat

cracking than chromium. During the PWHT (tempering), in the early stages,

coherent M02C type carbides precipitate first and the grain interiors become

harder (stronger). In the presence of other carbide formers like vanadium,

niobium, and titanium, that have a greater affinity for carbon, there is a

tendency to form more stable carbides. Even in such circumstances,

molybdenum leads to a large amount of solid solution hardening. M02Cis also

the carbide that gives great resistance to creep deformation.

Tamaki19-24 attempted to determine the separate and combined effects of

chromium and molybdenum on reheat cracking. A modified implant test was

used during these studies. The test was employed to determine the minimum

stress that would cause fracture within a 20hr period, while post weld heat

treating the sample at 600°C. Cracking susceptibility was related to the

magnitude of the critical stress for rupture (crAw-crit). The lower the stress the

higher the susceptibility. The results from these studies are shown in Figure

2.6.

\ \71>q 2Vb4 21

~ \10 0'1 '3 ?'2..;2....l

Figure 2.7 shows the influence of the chromium and molybdenum content, for

different stress levels. Susceptibility to reheat cracking with changing alloying

content is greatest in these diagrams where the stress contours are closest

together.

The plotted data can be divided into four regions:

• I : Alloy < 1% Cr and < 0.5% Mo. Relatively insensitive to cracking.

• IIa: Alloy 0-1% Cr and 0.5-1% Mo. Sensitivity increases with increasing

chromium and molybdenum content.

• lIb: Alloy> 2% Cr and 0.5-1 % Mo. Sensitivity decreases with increasing

chromium content.

• III: Alloy ± 1% Cr and> 1% Mo. Highest sensitivity to cracking.,

...•; 0...•c:g "

O.S 1.0 1.5Molybdenu~ content • wt'

Figure 2.7: Contour lines of critical restraint stress (crAW-c:r1t) shown on the Cr-Mocontent diagram 19.

A study on the carbides in the alloy system was also undertaken in order to

better understand the microstructural causes for differences in reheat cracking

susceptibility.2oThe results showed that materials with the highest fraction of

M2Ctype carbides, had the greatest tendency to crack. With a smaller amount

of M2C (or larger amount of M7C3 or M23C6), the susceptibility to reheat

cracking decreased. Figure 2.8 (a) and (b) depict these results graphically.

f:::It,.80 20 1600°c 24 hrl

rro IU I

2e::l..-\Eo\.l.c 1u

weightfractionof M2C

1.5wt\

4-----,------r------,,","" weight

(50) f=aC<:lon/ ,.,,-¥u ,,-

,I' . ..- ,-:,.. I

/ ..,0\ I

E~=1Eo'"""u

oo

Figure 2.8: (a) Carbides present in Cr-Mo steels tempered at 600°C for 24 h. oM3C,/

-MC2, AM7C3, AM23CS' (b) Weight fraction of M2C shown on the O"AW-crlt diagram 23.

Tamaki also undertook a study on secondary hardening and high t~mperature

hardness, because M2C and M7C3 strengthen the matrix by means of

precipitation. The samples were held at temperature for one hour both at the

holding temperature and at room temperature after cooling. At room

temperature the samples showed an increase in hardness, while at elevated

temperature the precipitation of the carbides resulted only in a delay in

softening (softening continues to occur but at a lower rate than that which

occurs at lower temperatures). The results of these tests are shown in Figure

2.9.

4ftIII

~ 300"0...III~

Ol .2 SO•..~...:lJ...;>

as-quenched

Hardness at roomtemperature

I,

Figure 2.9: Hot hardness and hardness at room temperature versus temperingtemperature23.

When the principal precipitate in the structure was M2C,the delay in softening

occurred at a higher temperature than when the largest fraction ~f precipitates;'

was M7C3. Tamaki postulated that the grain boundary embrittlement would be

of a similar nature in either type of alloy and therefore embrittlement of the

grain boundaries would initiate at the same temperature and proceed in

similar fashion for both types of alloys. This is shown schematically in Figure

2.10. According to this postulation a delay in softening of the matrix which

occurs at higher temperatures (large fractions of M2C carbide), the embrittling

of the grain boundaries may cause the intercrystalline flow stress to be less

than the intracrystalline flow stress and consequently intercrystalline fracture

can occur. If a delay in softening occurs at lower temperatures (large fraction

of M7C3 carbide precipitates), then the intracrystalline flow stress remains

lower than the intercrystalline flow stress over the whole temperature range

and intercrystalline fracture does not occur (Figure 2.10 (b)).

'Int racry's tall ineflow stress in HAZ

" lntercrystalline" flow st.ess in HAZ

"'~ Embrit:~ling the" grain boundary

"tJ, i\~.>l lntercrysealline.. ' '\"'" fracture

./1':"'< / \1Delay of softening' I I.at higher cemperat.ure i !

.T2! IT2-

·I·IIsoft.eningtemperat.ure

I·Delay ofat lower

Vanadium is added to Cr-Mo steels, because vanadium additions dramatically

increase elevated temperature strength. This is accomplished by the fine

precipitates of V4C3throughout the matrix. When considering the ~G and PSR

parameters vanadium unfortunately increases the susceptibility to reheat

Crackingdrastically (multipliers for vanadium are the greatest).

The fine V4C3precipitates throughout the matrix cause matrix strengthening.

This in turn leads to the accumulation of strain in the 'weaker' grain

boundaries25. It has been reported that early in the heat treatment cycle,

intense V4C3precipitation takes place at the ferrite-bainite interfaces due to

segregation effects. At temperatures of 500-550°C, coherent precipitation

occurs in the ferrite lattice similar to M02C precipitation. This leads to the

maximum hardness and strength.

A few researchers studied the vanadium-to-carbon ratio. Jones26 studied

welds in 1Cr-%Mo-%V materials and found that a vanadium-to-carbon ratio of

3.5-4.5 showed a high susceptibility to cracking. Stone and Murra~7 noted

that vanadium-to-carbon ratios of 3-4 induced a minimum in creep ductility.

When the ratio was reduced to 1.5 it increased ductility. Thus, a vanadium-to-

carbon ratio of 1.5-2 was recommended to mitigate reheat cracking.

Results obtained by Tamaki24are shown i~ Figure 2.11. It is apparent that

even small additions of vanadium reduce the critical restraint stress, where

the maximum effect is found in low chromium, low to high molybdenum alloys.

. ..1 2 0 1

content wt\!

::: 700~-. ..- --

300~. _,":-0. 3\V .----,--_ ....-

2 0Chromium

Figure 2.11: Effect of vanadium additions on the critical restraint stress, O"AW-crlt of Cr-Mo steels 24.

The increase in cracking susceptibility was said to be related to a decrease in

the rate-of stress relaxation in a similar fashion to that experienced in the Cr-

Mo alloys mentioned in the previous section. It has been suggested that the

decrease in the rate of stress relaxation due to a vanadium addition is brought

about by the precipitation of V4C3in addition to M02C.

Carbon obviously plays an important role in elevated temperature steels,

since carbide precipitation is involved in elevated temperature phenomena. In

the beginning the researchers overlooked the effect of carbon on reheat

cracking susceptibility, since it does not appear in the ilG or PSR parameters.

Some years later, carbon was included in the ilGj parameter and the multiplier

for carbon in that parameter is quite large. Ito and Nakanichi1also studied the

effect of carbon in alloys with two vanadium levels. They found that cracking

increased as carbon was increased from 0.05-0.10%, but was not changed by

a further increase to 0.25%. Other work showed that an increase in carbon,content also resulted in larger amounts of segregation of phosphorus at thegrain boundaries.28,29

Nickel does not appear in the ilG or PSR parameters and appears to have no

effect on cracking susceptibility. This was due to the small additions of Nickel.

In subsequent research work nickel contents above 1.5% increased thecracking susceptibilty.21

A high manganese to silicon ratio was found to decrease susceptibility to

reheat cracking in 2Y4Cr-1Mo SA welds.3oAn increase in manganese also

showed lower susceptibility in CrMoV steels.25Other researchers state that

manganese cosegregates with phosphorus to prior austenite grain boundariesand can act as a potent embrittling element,31,32

According to Ratliff and Brown33an increase in silicon appears to enhance the

rate of cementite dissolution and thereby promotes the precipitation of

chromium, molybdenum and vanadium carbides. This results in increased

secondary hardening. It appears that silicon accelerates the precipitation of

the M02Ccarbide the most.

Harris stated that while small amounts of titanium might appear beneficial (as

deoxidiser or grain refiner), large amounts as deliberate alloying additions can

increase cracking susceptibility due to matrix strengthening.34

,Phosphorus is the most potent, among the residual elements, in embrittling

the grain boundaries and thereby increases the susceptibility of the alloy to

reheat cracking. By adding the alloying elements chromium and molybdenum,

it was found that the solUbility of phosphorus in both ferrite and austenite

decreased markedly.30Segregation of phosphorus to the grain boundaries is

thus enhanced by chromium and molybdenum. Phosphide precipitation may

even occur at the grain boundaries.

When molybdenum is in solid solution, it ties up phosphorus in Mo-P clusters,

thus preventing phosphorus segregation to the grain boundaries.

Subsequently, during PWHT, Mo2Ccarbides are formed and the phosphorus

is released and segregation can take place. It has been suggested that the

change in phosphorus concentration in lerrite is closely related to the

formation of M2Ccarbides.

Wilkenson et al.35 have suggested that phosphorus, when segregated to prior

austenite grain boundaries, enhances nucleation of cavities.

Several notable results were obtained by Tamaki in his study of the effect of~

phosphorus in Cr-Mo steels :20,22,29,36

1. Plots were made for alloys low in phosphorus (0.010-0.013%) and high

in phosphorus (0.016-0.020%). It showed that increased phosphorus

levels moved the critical stress curves to lower molybdenum contents.

2. Phosphorus had already segregated to some extent in the HAZ during

the weld thermal cycle. This has to do with the ferrite to austenite

phase transformation. The welding cycle thus initiates the segregation

and embrittling process.

3. For a particular alloy, there exists a critical phosphorus level below

which embrittlement is not apparent. The minimum critical value of

phosphorus of 0.008% coincides with the 1Cr-%Mo composition.

4. For the range of alloys tested (0-2%Cr, 0.5%Mo), as long as the

phosphorus content was below a critical value for a particular alloy, the

critical stress level remained approximately the same.

I

Auger analysis has shown that sulphur also segregates to the fracture

surfaces. Sulphur segregation is associated with dislocation tangles along the

boundaries generated by impurity penetration. A grain boundary in this state

will be conducive to sliding and will inhibit migration. Cavity nucleation at

suitable particles may become possible. Sulphur in the form of grain boundary

sulphides has also been linked to the initiation of cavitation on the grainboundaries.37,38

Stress-induced segregation of sulphur to the grain boundaries, has also been

suggested.39,4o The sulphur segregates to the grain boundaries, by means of

grain boundary diffusion, ahead of the crack tip region during stress relief.

This in turn would lead to low crack growth energy requirements.

The most obvious way to avoid reheat cracking is to avoid using susceptible

materials. This in practice is not always easily conceivable and some changes

to the manufacturing/welding process may be beneficial:

• Use of a low-strength weld metal with high-strength base metals to allow

deformation to occur in the fusion zone rather than in the HAZ.

• Increasing heat input or preheat temperature. Higher energy inputs and

preheating result in slower cooling rates and softer transformation

products.

• Multiple-pass welds. This increases the amount of grain refinement due to

overlapping of the HAZ's. There is a limit on the amount of grain

refinement one wants to achieve as a too fine a grain size will reduce the

creep properties of certain components during operation. This is frequently

encountered during remnant life estimation where in-service creep

damage first appears and accumulates in the fine grained heat affected

zone (FGHAZ).

• Temper beads. In essence it is the same as multi-pass welds, where a,final run is made over the last weld bead in order to refine the

microstructure of the HAZ. It was found to be very beneficial.

• Complete austenitization after welding or normalising has been found to

increase the HAZ toughness and prevent cracking.

• Weld dressing to remove any surface discontinuities at the weld toes

effectively reduces cracking by eliminating crack initiation sites.

• Peening, to remove some of the residual tensile stresses at the surface of

the weld. Other processes to reduce the weld stresses such as

backstepping, interstage PWHT and simplifying the design and geometry

of the weldments can also be employed.

1. Ito, Y. and Nakanishi, M., "Study of Stress Relief Cracking in Welding Low Alloy Steels",

The Sumitomo Search, No.7, May 1972, pp. 27-36.

2. Easterling, K., "Introduction to the Physical Metallurgy of Welding", Butterworths, Boston,

1983

3. Macnerauch, E. and Wohlfahrt, H., "Different sources of residual stress as a result of

welding", Conference on Residual Stresses in Welded Construction and Their Effects",

pp. 267, The Welding Institute, London, November 1977.

4. American Welding Society, Recommended Practices for Welding Cr-Mo Steels,

ANSI/AWS 0.10.8-86, AWS, Approved November 18,1995.

5. Baker, R.G. and Nutting, J., "The Tempering of 2~Cr-1 Mo Steel After Quenching and

Normalizing", Journal of Iron and Steel Institute, July 1959, pp. 257-267.

6. Lundin, C.D., Henning, J.A., Menon, R and Khan, K.K., ''Transformation, Metallurgical

Response and Behaviour of the Weld Fusion Zone in Cr-Mo Steels for Fossil Energy

Application", Final Technical Report No. ORNUsub.81-7685/02&77, September 1987.

7. Buchmayr, B., Cerjak, H. and Fauland, H.P., ''The Effect of the Precipitation Behaviour on

the HAZ-Properties of 1Cr-Mo-V Steel", 2nd International Conference on Trends in

Welding Research, Gatlinburg, Tennessee, May 1989.

8. Boniszewski, T. and Eaton, N.F., " Electron Fractography of Weld Reheat Cracking in

CrMoV Steel", Metal Science, Vol. 3,1969, pp. 103-110.

9. Vinckier, A and Dhooge, A, "Reheat Cracking in Welded Structures during Stress-Relief

Heat Treatments", Journal of Heat Treating, Vol. 1(1), ASM, 1979, pp. 72-80.

10. Dhooge, A., Dolby, RE., Sebille, J., Steinmetz, Rand Vinckier, AG., "A Review of Work

Related to Reheat Cracking in Nuclear Reactor Pressure Vessel Steels", International

Journal of Pressure Vessels and Piping, Vo1.6(5),September 1978, pp. 329-409.

11. Presser, RI., McPherson, R, ''The Role of Trace Elements in Reheat Cracking",

Conference on Residual Stresses in Welded Construction and Their Effects", pp. 387-

398, The Welding Institute, London, November 1977.

12. Hipsley, CA, King, J.E. and Knott, J.F., "Toughness Variations and Intergranular

Fracture during the Tempering of Simulated HAZ Structures in a Mn-Mo-Ni Steel",

Proceedings of International Conference on Advances in the Physical Metallurgy and

Application of Steels, University of Liverpool, 21-24 September, 1981, Metals Society,

London, England, pp. 147-155.

13. Meitzner, C.F. and Pense, AW., "Stress-Relief Cracking in Low Alloy Steel Weldments',

Welding Journal, Research Supplement, VoI.48(10), 1969, pp. 431s-440s.

14. Nakamura, H., Naiki, T. and Okabayaski, H., "Fracture in the Process of Stress-

Relaxation Under Constant Strain", 1st International Conference on Fracture, September

1965, Vo12.Sendai, Japan, pp. 863-878.

15. Pense, AW., Gaida, e.J, Powell, G.T., "Stress Relief Cracking in Pressure Vessel

Steels", Welding Journal, Research Supplement, August 1971, pp. 374s-378s.

16. McMahon, C.J. Jr., Dobbs, RJ. and Gentner, D.H., "Stress Relief Cracking in MnMoNi

and MnMoNiCr Pressure Vessel Steels", Materials Science and Engineering, Vol. 37,

1979, pp. 179-186.

17. Boniszewski, T., "Reheat Cracking in 2Y-.Cr-1MoSA Weld Metal", Metal Construction,

14(9), 1982, pp. 495-496.

18. BrusGato, R, ''Temper Embrittlement and Creep Embrittlement of 2Y-.Cr-1Mo Shielded

Metal Arc Welding Deposits", Welding Journal, Research Supplement, April 1970, pp.

148s-156s.

19. Tamaki, K., Suzuki, J., "Effect of Chromium and Molybdenum on Reheat Cracking

Sensitivity of Steels", Transactions on the Japan Welding Society, Vol. 14(2), October

1983, pp. 39-43.

20. Tamaki, K., Suzuki, J., "Assesment of Reheat Cracking Sensitivity of Steels from their

Chemical Composition", Metals/Materials Technical Series, 85 ASM's International

Welding Congress, Toronto, Canada, 14-17 October, 1985, pp. 1-8.

21. Tamaki, K., Suzuki, J., "Reheat Cracking Test on High Strength Steels by a Modified

Implant Tesf', Transactions on the Japan Welding Society, Vol. 14(2), October 1983, pp.

33-38.

22. Tamaki, K, Suzuki, J., "Assesment of the Reheat Cracking Sensitivity of Cr-Mo Steels",

The 4th International Symposium of the Japan Welding Society, November 1982, Osaka,

4JWS-II-17, pp. 467-472.

23. Tamaki, K, Suzuki, J., Nakaseko, Y. and Tajiri, M., "Effect of Carbides on Reheat

Cracking Sensitivity", Transactions of the Japan Welding Society, Vol. 15(1), April, 1984.

24. Tamaki, K, Suzuki, J. and Kojima, M., "Combined Influence of Chromium, Molybdenum

and Vanadium on Reheat Carcking of Steels", IIW Document IX-1518-88.

25. Mullery, F. and Cadman, RO.L., "Cracking of Welded Joints in Ferritic Heat-Resisting

Steels", British Welding Journal, April 1962, pp. 212-220.

26. Jones, KE., "Stress-Relief Cracking of Welded Pipe/Casting Joints in Cr-Mo-V Steels",

Paper 5 in Proceedings of Conference on Welding Creep Resistant Steels, The Welding

Institute, Cambridge, 17-18 February, 1970, pp. 66-78.

27. Stone, P.G. and Murray. J.D., "Creep Ductility of Cr-Mo-V Steels", Journal of Iron and

Steel Institute. Vol. 203, November, 1965, pp. 1094-1107.

28. Ehart, H., Grabke, H.J. and Onel, K, "Grain Boundary Segregation of Phosphorous in

Iron Base Alloys: Effects of Carbon, Chromium and Titanium", Proceedings of

Conference on Advances in Physical Metallurgy and Application of Steels, London, 21-24

September, 1981, pp. 282-285.

29. Tamaki, K, Suzuki, J., "Effect of Phosphorous on Reheat Cracking in Cr-Mo Steels",

ResearchReports of the Faculty of Engineering, Mie University, Vol. 9, December 1984,

pp.8-16.

30. Anonymous, "Testing Techniques to Study the Susceptibility to Reheat Cracking of

Carbon-Manganese and Low Alloy Steels", Welding in the World, Vol. 12(11/12), 1974.

31. Bodnar, RL., Ohhashi, T. and Jaffee, RI., "Effects of Mn, Si and Purity on the design of

3.5 NiCrMoV, 1CrMoV and 2.25Cr-1Mo Bainitic Alloys", Metallurgical Transactions A, Vol.

20A, August 1989, pp. 1445-1460.

32. Weng, Y.Q. and McMahon, C.J. Jr., "Interaction of Phosphorous, Carbon, Manganese

and Chromium in Intergranular Embrittlement of Iron", Materials Science and Technology,

Vol. 3, March 1987, pp. 2207-216.

33. Ratliff, J.L. and Brown, R.M., "The Deleterious Effect of Small Quantities of Aluminium on

the Stress-Rupture Properties of a Cr-Mo-V Steel", Transactions of the American Society

of Metals, Vol. 60, 1967, pp. 176-186.

34. Harris, P. and Jones, KE., "The Effect of Composition and Deoxidation Practice on the

Reheat Cracking Tendencies of Y2Cr-Y2Mo-XVSteel", Proceedings of International

Conference on Welding Research Related to Power Plants", England, 17-21 September,

1972, pp. 369-380.

35. Wilkenson, D.S., Abiko, K., Thyagarajan, N. and Pope, D.P., "Compositional effects on

the Creep Ductility of a Low Alloy Steel", Metallurgical Transactions A, Vol. 11A,

November 1980, pp. 1827-1836.

36. Tamaki, K., Suzuki, J., "Recent Studies on Reheat Cracking of Cr-Mo Steels",

Proceedings of International Conference on Stress Relieving Heat Treatments of Welded

Steel Constructions, 6-7 July, 1987, Bulgaria, pp. 325-326.

37. You, C.P., Hippsley, CA and Knott, J.F., "Stress Relief Cracking Phenomena in High

Strength Structural Steels", Metal Science, Vol. 18, August 1984, pp. 387-394.

38. Dolby, R.E., "Advances in the Welding Metallurgy of Steels", Proceedings of International

Conference on Advances in the Physical Metallurgy and Application of Steels, 21-24

September 1981, The Metals Society, London, pp. 111-125.

39. Shin, J. and McMahon, C.J. Jr., "Mechanisms of Stress-Relief Cracking in Ferritic Steel",

Acta Metallurgica, Vol. 32(9),1984, pp. 1535-1552.

40. Hippsley, CA, "Brittle Intergranular Fracture at Elevated Temperatures in Low Alloy

Steels", Materials Science and Technology, June 1985, Vol. 1, pp. 475-479.

41. Lundin, C.D. and Khan, K.K., "Fundamental Studies of the Metallurgical Causes and

Mitigation of Reheat Cracking in 1Y-.Cr-Y2Moand 2Y-.Cr-1MoSteels", WRC Bulletin 409,

February 1996.

To determine a certain steel grade's susceptibility towards reheat cracking a

suitable test method must be selected or developed. The specimen should if

possible incorporate a welded joint similar to the joint used in the components

like headers and main steam pipes. The joint should also be manufactured

with the same heat inputs, welding processes, consumables, pre-heating, etc.

Furthermore, the thermal treatment during testing should aim to duplicate as

closely as possible the prescribed PWHT thermal cycle.

When reviewing the many papers that studied the phenomenon of reheat,cracking it becomes quite clear that a multitude of tests exists and has been

employed. Comparing the test re~ults is complex and sometimes lead to

confusion. In order to simplify the overall number of tests, three categories

were proposed to classify the tests1:

A. Tests on complete weldments

1. Lehigh restraint test

2. Restrained fillet weld test

3. V-groove test

4. Stellite bead test

5. BWRA ring test

6. Steampipe butt weld test

7. H-type restraint test

8. Circular patch restraint test

9. Strained root bead test

10.Restrained butt weld test

11.MRT test

12.NRC Regulatory Guide cladding test

B. Tests on specimens containing a weld

1. Jigged stress relaxation test

2. Isothermal constant load rupture test

3. Implant test

4. Controlled heating rate, constant load test

5. Vinckier test - Stainless backing bar test

C. Tests on specimens containing a thermally simulated HAZ

1. Isothermal stress relaxation test

2. Isothermal constant load rupture test

3. Isothermal slow strain-rate tensile test

4. Controlled heating rate stress relaxation test

5. Controlled heating rate constant load test

6. Stainless sleeve test

7. Pre-cracked bend test

8. Slit tube test

9. Jigged stress relaxation test

I

The use of tests on complete weldments has the advantage of being related

to actual weld and PWHT conditions. It has been found however that

reproducibility is a significant problem. Results show considerable scatter with

one sample cracking and another showing no cracking. The complex

interactions of residual stress and strain, their distribution and magnitude,

relaxation and triaxiality, together with progressively changing microstructures

during PWHT are difficult to incorporate in a small weld sample. A small test

specimen will also experience a significantly smaller amount of creep strain

than a large welded structure.

Weld simulation tests have the advantage of reproducibility, known stress

level and control of the microstructure in the test area. It is also possible to

accurately locate or place a notch in a /well-defined microstructure. The

disadvantages are that strains associated with weld contractional stresses are

not duplicated and in general only one area of the HAZ is tested and the

effects of the adjacent weld metal and base metal are not present. .

The existing tests mentioned above all suffer from one or more of the

following disadvantages:

1. Difficulty in quantifying susceptibility

2. Poor correlation with field experience

3. Poor reproducibility

4. Lack of one or more of the factors necessary for reheat cracking

5. Requirement of expensive instrumentation or elaborate testing facility

6. Only one region of the HAZ is tested and the effects of weld metal, other

HAZ regions and base metal are not accounted for in the test

To overcome the deficiencies of these tests a test method developed by

Lundin1 was used in this program for studying the susceptibility of service-

exposed material towards reheat cracking. This test has the capability of

simultaneous evaluation of reheat cracking in various regions of the

weldment. It also incorporates all the necessary factors for reheat cracking to

occur as listed in the literature survey as summarised by Ito and Nakanishi2.

,The test is called the 'Spiral Notch Test'. The use of a notch is important, as

many researchers have noted th~t the effect of a stress raiser is very

important in the mechanism of reheat cracking. The sample is exposed to a

simulated weld thermal cycle with a peak temperature (TM) of 1300°C in the

centre of the sample. Lower temperature transformation products as well as

the unaffected base metal are also incorporated into the test length of the

sample. A schematic diagram of such a sample is given in Figure 3.1. It can

be seen from the figure that every metallurgically different region of the weld

is notched similarly and thus the stress concentration experienced by every

region will be virtually identical because the notch extends through the base

metal and weld HAZ. The sample is then inserted into a creep frame at a

constant load. A protective gas atmosphere protects the sample from

excessive oxidation during testing. The gas consisted of 5% Hydrogen (H2)

and the balance Argon (Ar). The region and mode of fracture for a specific

stress, during the PWHT thermal cycle can then be determined. A schematic

representation of the test procedure is shown in Figure 3.2. It was decided to

conduct isochronous rather than isothermal tests. Thus the full effect of the- -PWHT thermal cycle could be evaluated as well as the specific failure

temperature for a given load applied. Due to time constraints it was decided to

use a heating rate of 400°C/h for the first hour up to 400°C and afterwards

100°C/h until failure. The peak temperature of the PWHT was set at 750°C.

Peak temperature incenter of specimen:1300°C

Temperaturecycle

Applied stress(constant) .

Figure 3.2: Schematic representation of thermal and stress cycles for reheat crackingstudies

CGHAZSimulation

The materials received were analysed to confirm the grade. Eskom accredited

methods 106 (alloying elements) and 119 (carbon) were used. The tests were

conducted at Technology Services International (TSI). TSI is company that

forms part of the Eskom group.

As indicated, the spiral notch test method makes use thermally simulated test

samples. A peak temperature of 1300°C was selected as this closely relates

the peak temperature during welding for the CGHAZ. The weld simulator

heats the sample by means of electrical current flowing through the sample.

An S-type thermocouple in conjunction with a REX P-200 controller unit

regulates the thermal cycle. Due to the size of the sample and the limitation of

the simulator the heating rates of actual welding could not be replicated

accurately. Heating rates of around 250°C/s were achievable. To more closely

simulate actual welding conditions with a heat input of 1kJ/mm, it was decided

to control the cooling down time from 800°C to 500°C (~t8-5).This region is

critical for the type of transformation products that will form when welding is

performed. When ~t8-5 is properly controlled the effect of pre-heating as

described in the welding procedure specification (WPS) is taken into account.

Thus it is possible to closely simulate actual welding conditions. As an

example the conditions during simulation of the X20 material is shown.I

~t = q/v (3-1)8-5 21Z"A0

1

1 1 1 (3-2)=

°1773-To 1073 -To

q/v = heat input (kJ/mm)

')..= Thermal conductivity (Jm-1s-1K-1)

Equations 3-13 and 3-23 are used to calculate the amount of time that it will

take to cool down from 800°C to 500°C for a certain preheat temperature. For

X20 at a preheat or interpass temperature of 300°C (A. = 26 Jm-1s-1K-1)a .:1.t8-5

of 18 s was calculated. During actual simulation the recorded time was 16 s.

Thus for all four grades of steel tested the simulated HAZ represented the

HAZ of welded material with a heat input of 1kJ/mm. The total length of the

HAZ was 20 mm.

The CGHAZ simulated samples were then machined according the

dimensions as shown in Figure 3.1.

Due to the fact that no service-exposed P91 material was available from

power plants, it was decided to artificially age some of the available newmaterial.

Blank samples for CGHAZ simulation were placed in a stainless steel tube

with some zirconium chips. The tube was welded shut and heated in a

furnace at 650°C for around 860h. This corresponds to a service exposure at

550°C of more than 150 000 hours.

/3.5 INSTRUMENTS AND MEASUREMENTS DURING TESTING

The creep frame used was a Denison machine, model T.48. It has a capacity

of 1000 kg. A REX P-250 controller and three K-type thermocouples

controlled the furnace.

The knife-edges visible on the sample were used to attach the elongation rig

of the creep frame. Two LVDT's (Linear Variable Differential Transducers)

were attached to the elongation rig. The displacement of the LVDT's was

used to measure the elongation of the sample during testing. The type of

LVDT used was Schaevitz PCA 116-300.

A K-type thermocouple was spot welded to the sample during testing. Thus

the thermal cycle and failure temperature experienced by the sample could be

accurately recorded.

The sample temperature and elongation was measured and logged using a

computer with a data logging card as well as a Yokagawa HR 1300 Recorder.

One half of a fractured sample was used for metallography and the other for

fractography. The fracture surfaces were studied by scanning electron

microscopy (SEM), to assess the fracture mechanism. The fracture surfaces

are represented in the following sections for each steel grade.

Certain samples were taken to study the microstructure of the four different

steel grades, in the new and service exposed condition, as well as in the as-

simulated and after testing (PWHT) condition. The samples were sectioned

longitudinally so that a full transverse view ranging from the HAZ to base

metal was possible. The samples were mounted and ground and polished to a/

1Jlm finish. The %Cr-%Mo-%V and 2%Cr-1Mo grades were etched with 2%

Nital. The X20 and P91 grades were etched with Vilella's etchant. The etched

samples were then studied using an optical microscope to determine the

various microstructures of the different regions in each sample. The different

microstructures are represented in the following sections for each steel grade.

The metallographic samples were used for the hardness survey. Vickers

microhardness with a load of 300g was taken. The results for the hardness

survey are represented in the following sections for each steel grade.

1. Lundin, C.D. and Khan, K.K., "Fundamental Studies of the Metallurgical Causes and

Mitigation of Reheat Cracking in 1~Cr-Y2Mo and 2~Cr-1Mo Steels", WRC Bulletin 409,

February 1996.

2. Ito, Y. and Nakanishi, M., "Study of Stress Relief Cracking in Welding Low Alloy Steels",

The Sumitomo Search, No.7, May 1972, pp. 27-36.

3. Easterling, K., "Introduction to the Physical Metallurgy of Welding", Butterworths, Boston,

1983.

One of the most widely used creep resistant low alloy steels is %Cr-%Mo-1,4V.

The material is supplied in the tempered condition with a ferritic and pearlitic

structure. The presence of vanadium is the primary reason for the creep

strength of the material. Some of the more common properties of this grade

are listed below1• The service-exposed material was obtained from Hendrina

power station and operated at 540°C for 150 000 hours.

ELEMENT min % max %C 0.10 0.18Si 0.10 0.35Mn 0.40 0.70P 0 0.035S 0 0.035Cr 0.30 0.60Mo 0.50 0.70V 0.22 0.32

PROPERTY VALUEYield stress >= 345 MPa

Tensile Strenqth 490 - 640 MPaElonqation >= 20 %

Impact value (DVM) / >= 41 JHardness Brinell 145 - 190 HB30

Mechanical Properties at ambient temperature (Values for longitudinal sample bars <=60mm 0

Eskom accredited methods No 106 (Alloying elements) and 119 (Carbon)

were used on the new and service-exposed material. The results are shown in

Table 4.1. The original result sheets are attached in Appendix 1.

Element New Material Service-exposed Material(± 150 000 h at 540°C)

Carbon 0.14 0.19Chromium 0.46 0.53

Nickel 0.08 0.16Manganese 0.51 0.54Molybdenum 0.51 0.54Vanadium 0.28 0.31

Sulphur 0.01 0.012Phosphorus 0.008 0.014

Silicon 0.12 0.24Titanium 0.01 0.01Copper 0.09 0.15Cobalt 0.01 0.02

Niobium 0.005 0.005Tin 0.01 . 0.02

Tungsten 0.005 0.005Table 4.1: Chemical analysis for new and service-exposed %Cr-%Mo-Y4V

material

The chemical composition of the test material was within specified range for~Cr-~Mo-~V material.

4.3 SPIRAL NOTCH TEST RESULTS (Constant load, PWHT Thermal

Cycle)

The results from the spiral notch tests on the new and service-exposed ~Cr-I

~Mo-~V are shown in Table 4.2 and Graphs 4.1 and 4.2.

Nominal Stress %Cr-%Mo-Y4V %Cr-%Mo-Y4V

(MPa) New Material Service-exposedMaterial

356 CGHAZ Base Metal305 CGHAZ Base Metal204 CGHAZ Base Metal153 - Base Metal

800

700

600

500

Ee2 400i!"a.E"...

300

200

100

0153

Graph 4.1: Temperature of final fracture for %Cr-%Mo-1j4V steel samples

20

18

16

14

_ 12Cco0

~ 10Clco0in

8

6

4

2

0153

The experimental graphs compiled during these tests are attached in

Appendix 2.

The fracture surfaces of the samples tested at 204 MPa for both new and

service-exposed material were studied in the SEM to confirm the mechanism

of fracture. The fracture surfaces are shown in Figures 4.1 to 4.5.

Figure 4.1: %Cr-%Mo-Y..v New material - Intergra~ular fracture surface, 204 MPa

Figure 4.2: High Magnification of Figure 4.1. Some microductility on prior austenitegrain boundaries

Figure 4.3: Intergranular fracture along prior austenite grain boundaries, area of higher

magnification than Figure 4.1

Metallographic samples were prepared from new and service-exposed

material after the CGHAZ thermal simulation. The samples tested at 204 MPa

(new and service-exposed) were also taken and metallographic samples were

prepared. The resulting microstructures are shown in Figures 4.6 to 4.13.

Descriptions of the structures are given in Table 4.3.

As-simulated

Service-exposed

Service-exposed

FigureNo.

4.6

4.10

4.11

4.12

4.13

Description

Base metal. Ferritic matrix with partially spheroidised pearlite. There is a dispersion of fine carbideson the rain boundaries. ASTM rain size of around 7.CGHAZ. Untempered bainitic microstructure. Might be upper bainite. FGHAZ will have the samestructure as Fi 4.7 with smaller rains.Base metal. Ferritic matrix with a dispersion of coarse and continuous carbides on the grainboundaries. Still some signs of the decomposed secondary phase (Pearlite). Ferrite has an ASTM

rain size of about 9.CGHAZ. Untempered bainitic microstructure. In essence the same as the condition of the CGHAZof the new material.Base metal. Ferritic matrix with partially spheroidised pearlite. There is a dispersion of fine carbideson the rain boundaries. ASTM rain size of around 7. Some areas that a ear overtem ered.Tern ered Bainite. Inside of the needles have broken_a art. Fine carbides throu hout the matrixBase metal. Ferritic matrix with a dispersion of coarse and continuous carbides on the grainboundaries. Still some si ns of the decom osed seconda hase Pearlite.CGHAZ. The structure consists of tern ered bainite.

Hardness profiles were determined from the same samples used for

metallography. A Vickers microhardness-testing machine with a load of 300g

was used. The profiles start at the centre of the HAZ and extend through to

the base metal. The profiles are shown in Graph 4.3.

i500..,e 450

= 400c'E~ 3502~ 300

I!!.: 250

5 200

--+- As-Simulated: New

--- As-Simulated:Service-+-PWHT(40kg):New

-*-PWHT 40k :Service

From the spiral notch tests it is quite clear that there is a distinct difference

between the behaviour of the new and service-exposed material. The new

material underwent less elongation than the service-exposed material and all

samples tested fractured in the CGHAZ area of the sample. All the service-

exposed samples fractured in the base metal area of the samples and at

lower temperatures than the corresponding new material.

The fracture surfaces studied under the SEM clearly shows two different

fracture mechanisms that lead to failure during testing. The new material

failed by a brittle, intergranular failure mechanism at all loads tested. The

service-exposed material all failed in a ductile manner at all loads tested.

Some microductility along the prior austenite grains boundaries of the new

material were observed. The fractography shows that the new material failed

due to a reheat cracking mechanism.

The state of the microstructures shows that the CGHAZ simulation for new

and service-exposed material delivers almost an identical HAZ in the material.

Grain sizes were larger than expected during actual welding but this is a

common occurrance when weld simulation is performed. After PWHT (spiral

notch test) the HAZ's of both states showed tempering of the bainite. Very,

minor changes were observed in the base metal areas. In the new and

service-exposed material, the carbides were taken up into solution during the

CGHAZ simulation. The cooling rate from the peak temperature is to high for

these carbides to re-precipitate thus they remain in solid solution. During the

subsequent PWHT these carbides re-precipitate adding strength to the matrix.

The base metal carbides did not go through significant changes during

PWHT. These carbides, in the new material, are still in the meta-stable state

that contributes to creep strength. The carbides in the service-exposed

material have already either mostly or all transformed to the more stable

carbide type with a lower creep strength.

The hardness profiles also showed that after CGHAZ simulation the hardness

for new and service-exposed materials are ~imilar in the HAZ. After PWHT an

increase in hardness was found in the HAZ; due to secondary hardening for

both the new and service-exposed material. The re-precipitation of carbides

took place during the PWHT. In general the hardness of the service-exposed

material is however lower than that of the new material.

It was found that %Cr-%Mo-'Y,.v material in the unexposed condition is

susceptible for reheat cracking after welding. Referring to Figure 4.2 the

microductility observed on the grain boundaries indicates reheat cracking but

that the mechanism is not wholly a brittle mechanism. Some precipitate free

or denuded zones in close proximity to the grain boundaries playa role in the

cracking mechanism.

The service-exposed material fractured in the base metal and was

consequently not susceptible to reheat cracking, for the test conditions used.

The carbides in the base metal of the service-exposed are in the stable state

and have lower creep strength. The softer base metal of the service-exposed

state protected the HAZ during PWHT by being the primary area to undergo

deformation under the influence of tensile stress due to the fact that there is

little or no meta-stable carbides that add creep strength. The starting,

microstructure of the material before welding apparently plays a significant

role in the susceptibility of the material with regards to reheat cracking.

Another commonly Cr-Mo creep resistant alloy used is 2~Cr-1 Mo. Up to the

1950's it was highest ferritic alloy used in boiler applications. The material is

supplied in the quenched and tempered condition with a ferritic and bainitic

structure. Some of the more common properties of this grade are listed

below 1. The service-exposed material was obtained from Kriel power station

and operated at 510°C for 130 000 hours.

ELEMENT min% max%C 0.08 0.14Si 0 0.50

Mn 0.40 0.80P 0 0.030S 0 0.025Cr 2.00 2.50Mo 0.90 1.10Cu 0 0.30


Tensile Strength 480 - 660 MPaElonQation >= 18 %

Impact value (DVM) ! >= 55 JHardness Brinell 130 -175 HB30

Mechanical Properties at ambient temperature (Values for longitudinal sample bars <=60mm 0




Element New Material Service-exposed Material(± 130 000 h at 510°C)

Carbon 0.11 0.10Chromium 2.41 2.57

Nickel 0.05 0.17Manganese 0.49 0.54Molybdenum 0.87 0.91

Vanadium 0.01 0.02Sulphur 0.001 0.003

Phosphorus 0.018 0.021Silicon 0.11 0.04

Titanium 0.01 0.01Copper 0.04 0.12Cobalt 0.01 0.03

Niobium 0.005 0.005Tin 0.01 . 0.01

Tungsten 0.005 0.02Table 5.1: Chemical analysis for new and service-exposed 2Y4Cr-1Mo

material

The chemical composition of the test material was within specified range for

2~Cr-1 Mo material.


Cycle)

The results from the spiral notch tests on the new and service-exposed 2~Cr-. ;/

1Mo are listed in Table 5.2 and Graphs 5.1 and 5.2.

Nominal Stress 2Y4Cr-1Mo 2%Cr-1Mo


356 Base Metal Base Metal305 Base Metal Base Metal204 CGHAZ Base Metal153 Base Metal Base Metal

800

700

600

500

Ee"f 400..Q.

E.....300

200

100

0153

_ 12~co~ 10OlcoW

Graph 5.2: Amount of elongation for 2Y~Cr·1Mo steel samples. -


Appendix 2.


service-exposed material was studied in the SEM to confirm the mechanism

of fracture. The fracture surfaces are shown in Figures 5.1 to 5.5.

Figure 5.2: High Magnification of Figure 5.1. Some microductility on prior austenitegrain boundaries

Metallographic samples were prepared from new and service-exposed

material after the CGHAZ simulation. The samples tested at 204 MPa (new

and service-exposed) were also taken and metallographic samples were



Figure 5.7: CGHAZ of new material- 200x (After thermal simulation)

~

Figure 5.11: CGHAZ of new material - 500x. (After PWHT). Secondary intergranular

crack visible

As-simulated

Service-exposed

Service-exposed

FigureNo.

5.6

5.11

5.12

Description

Base Metal: Ferritic matrix coupled with partly decomposed tempered Bainite. There is also adis ersion of fine carbides on the rain boundaries. ASTM rain size of a roximatel 8.CGHAZ: Bainite. In the FGHAZ the structure is the same although grains are smaller. The FGHAZa eared to be more tern ered than The CGHAZ.Base Metal: No ferritic matrix left. Tempered, partly decomposed Bainite with a dispersion of finecarbides throu hout the structure.CGHAZ: Bainite. In the FGHAZ the structure is the same although grains are smaller. The FGHAZa eared to be more tern ered than The CGHAZ.Base Metal: Ferritic matrix coupled with partly decomposed tempered Bainite. There is also adispersion of fine carbides on the grain boundaries. The amount of carbides present seem to be

reater that in the As-simulated condition.CGHAZ: Partl tern ered Bainite with a dis ersion of fine carbides throu hout.Base Metal: Islands of ferrite are visible. Coarse carbides present on the grain boundaries.Tern ered artl decom osed Bainite with a dis ersion of fine carbides is also resent.CGHAZ: Tempered Bainite (needles broken apart). There is also a dispersion of fine carbides onthe rior austenite boundaries.

Hardness profiles were determined from the same samples used in the




j500...•s!. 450

i 400'l!~ 350ei 300

!!.= 250u:> 200

-+- As-Simulated: New

-- As-Simulated: Service-'-PWHT(40kg): New

--- PWHT 40k : Service

Graph 5.3: Hardness profile of 2Y4Cr-1Mo material - New and Service

exposed

From the spiral notch tests only one sample of the new material, tested at 204

MPa, failed in the CGHAZ. The rest of the samples all failed in the base metal

area. 1he service-exposed material all failed in the base metal region, for all

loads applied during the spiral notch test. For all loads applied the amount of

elongation was higher for new material in comparison with service-exposed

material, except for the tests conducted at 204 MPa, where the new material

failed in a brittle manner in the CGHAZ, with lower elongation than the

service-exposed material. Failure of the service-exposed material took place

at lower temperatures than the comparative new material for all loads applied.

The fracture surfaces studied by SEM clearly show two different fracture

mechanisms that lead to failure during testing (204 MPa new and service-

exposed). The new material failed with a brittle, intergranular failure

mechanism. The service-exposed material all failed in a ductile manner at all

loads tested. Some microductility along the prior austenite grains boundaries

were observed. The fractography shows that the new material, tested at 204

MPa failed due to a reheat cracking mechanism.

I

The state of the microstructures shows that the CGHAZ simulation for new

and service-exposed material resulted in almost an identical HAZ in the

material. Grain sizes were larger than expected during actual welding but this

is common when weld simulation is performed. After PWHT (spiral notch test)

the HAZ's of both states showed tempering of the bainite. A real difference

was observed in the base metal microstructure of the two different states of

2~Cr-1 Mo. For new material it was ferrite and tempered bainite. For the

service-exposed material no ferritic structure was present; it only consisted of

decomposed bainite with carbides dispersed throughout the matrix. The creep

properties of the service-exposed material were thus lower than that of the

new material during the PWHT (spiral notch test). This is due to the fact that

the structure is decomposed and that the meta-stable carbides that add creep

strength have transformed to the stable carbide type, with a lower creep/

strength.

The hardness profiles also showed that after CGHAZ simulation, the hardness

for new and service-exposed materials are fairly similar in the HAZ. The base

metal hardness of the service-exposed material was mostly lower than that of

the new material. No significant hardening of the HAZ took place during

PWHT, probably due to the lack of vanadium. No significant changes took

place in the base metal of the new and service-exposed material during

PWHT.

Unexposed (new) 21,4Cr-1Mo material can be susceptible for reheat cracking

after welding, though not as severely as new }'2Cr-}'2Mo-~V. The lack of

vanadium as alloying element probably leads to 2~Cr-1 Mo material being

less susceptible. No severe secondary hardening can take place, only chrome

and molybdenum carbides add to creep strength. The service-exposed

material showed no susceptibility for reheat cracking, for the test conditions

employed. The softer base metal of the service-exposed state protected the

HAZ during PWHT by being the primary area to undergo deformation under

the influence of tensile stress due to the fact that there is little or no meta-I

stable carbides that add creep strength. It seems that starting microstructure

(state of the carbides) before welding can playa role with regards to reheat

cracking.

X20CrMoV12-1 (X20) material was developed during the sixties and

seventies to enhance the properties of creep resistant steels for elevated

temperature use to improve boiler efficiency. In South African power stations

the material was first used on wide scale during the late 1970's and during the

1980's. The steel is normally used in the quenched and tempered condition

and the starting microstructure is tempered martensite. The service-exposed

material was obtained from Duvha power station and operated at 540°C for

150000 hours.

6.1.1 General Chemical Composition 1

ELEMENT min% max%C 0.17 0.23Si 0 0.50Mn 0 1.00P 0 0.030S 0 0.030Cr 10.0 12.5Mo 0.80 1.20Ni 0.30 0.80V 0.25 0.35

6.1.2 Mechanical Properties 1


Tensile Strength 690 - 830 MPaElonQation >= 16 %

Impact value (DVM) >= 41 JHardness Brinell 205 - 250 Hfho

Mechanical Properties at ambient temperature (Values for longitudinal sample bars <=60mm 0.

6.2 CHEMICAL ANALYSIS


were used on the new and service exposed material. The results are shown in


Element New Material Service Exposed Material(± 150 000 h at 540°C)

Carbon 0.22 0.19Chromium 12.05 12.29

Nickel 0.75 0.48Manganese 0.59 0.45Molybdenum 0.82 0.78

Vanadium 0.26 0.30Sulphur 0.004 0.013

Phosphorus 0.016 0.020Silicon 0.07 0.013

Titanium 0.01 0.01Copper 0.11 0.09Cobalt 0.01 0.01

Niobium 0.01 I 0.005Tin 0.01 0.01

TunQsten 0.005 0.005Table 6.1: Chemical analysis for new and service exposed X20CrMoV121

material

The chemical composition of the test material was within specified range forX20 material.


Cycle)

The results from the spiral notch tests on the new and service exposed X20

are listed in Table 6.2 and Graphs 6.1 and 6.2.;/

Nominal Stress X20 X20


407 Base Metal Base Metal356 Base Metal Base Metal305 Base Metal Base Metal254 Base Metal Base Metal

800

700

600

500Ee~ 400..Co

E.....300

200

100

0254

_ 12~co1ii 10OlcoW


Appendix 2.


service exposed material was studied by SEM to confirm the mechanism of

fracture. The fracture surfaces are shown in Figures 6.1 to 6.4.

Metallographic samples were prepared from new and service exposed

material after the CGHAZ thermal simulation. The samples tested at 254 MPa

(new and service exposed) were also taken and metallographic samples were



!

Condition State Figure DescriptionNo.

6.5 Base Metal: Tempered martensite (quenched and tempered condition), with a dispersion of fine

New carbides on the prior austenite grain boundaries, visible as strings.CGHAZ: Partly tempered martensite. Fine carbides are visible in some of the scattered tempered6.6

As-colonies.Base Metal: Tempered martensite (quenched and tempered condition), with a dispersion of finesimulated

Service 6.7 carbides on the prior austenite grain boundaries, visible as strings. Some b-ferrite also present.

exposed Some Qrains also show the ferrite with carbides (these areas are normally softer).CGHAZ: Fine carbides dispersed throughout the structure. Structure looks tempered - probably6.8 due to auto temperinq during the CGHAZ simulation.Base Metal: Tempered martensite (quenched and tempered condition), with a dispersion of fine

6.9 carbides on the prior austenite grain boundaries, visible as strings. Overall the structure is moretempered than the structure in the as-simulated condjtion.

PWHT New

6.10 CGHAZ: Partly tempered martensite. Some martensite still appears fresh while other martensiteareas are more tempered with fine carbides.





'6i 500.....2. 450::!! 400'E!.l! 350l!i 300

j 250

~ 200

The results from the spiral notch test revealed a significant difference between

the low alloy ferritic grades (%Cr-%Mo-1,.4Vand 21,.4Cr-1Mo) an.d a higher

alloyed grade like X20 (12Cr-1Mo). All the samples tested, both new and

service.:exposed material failed in the base metal area of the sample at all the

different testing loads. The failure temperature of the new material for all the

samples tested were higher than the comparative service-exposed material.

The amount of elongation per sample was higher for the new material than

the service-exposed material. The ductility of X20 at elevated temperature is

superior to the low alloy ferritic grades. The strength of the material is due to

the high amount of chromium that is present in the structure as chromium

carbides as well as molybdenum carbides.

From the SEM photographs it is evident that both the new and service-

exposed material failed due to a ductile fracture mechanism. Due to the notch

a stress concentration is present and the yield strength of the base metal is

lower than the HAZ during the PWHT. The base metal will be the primary

region where creep deformation occurs under the influence of a tensile stress

up to the point of failure.

The structure of the base metal in the service-exposed condition is such that it,

will be softer than the base metal of the new material. This leads to the failure

in the base metal areas because the yield strength for the base metal is lower

than that of the CGHAZ. CGHAZ simulation resulted in similar structures of

partly tempered martensite. This is due to the very high hardenability of X20.

A fUlly martensitic structure can thus be obtained in these materials with

slower cooling rates in comparison with that of to the two ferritic grades (%Cr-

%Mo-~V and 2~Cr-1Mo). During cooling of materials with a high

hardenability the martensite that formed first is still at a relatively high

temperature and longer periods of time and undergoes auto-tempering as the

work piece cools down to room temperature. The structure of the sample that

underwent PWHT (spiral notch test) showed a tempered structure due to the

thermal cycle as expected.

The hardness profiles showed the service-exposed material base metal had a

lower hardness than the new material. The hardness of the HAZ's showed

some difference. After PWHT the hardness of the HAZ is significantly reduced

as toughness and ductility are restored due to the tempering of the hardened

structure. The hardness (yield strength) is still higher than that of the base

metal, which points to the fact that the base metal is the area where creep

deformation was concentrated during PWHT.

X20 material creep strength is obtained due to the high content of chromium

as alloying element with molybdenum in small quantities. The hardenability of

the steel is very high. PWHT restores toughness and ductility and does not

seem to increase the strength, as is the case with ~Cr-~Mo-~V. Due to the

high hardenability, a X20 component during welding operations is held at a

certain temperature before PWHT. This could prevent full transformation of

the HAZ to martensite and retained austenite could be present before PWHT.

The temperature before PWHT must be such that it prevents cracking due to

the high hardenability and also to allow almost full transformation. The yield,

strength of the base metal for new and service-exposed material will always

be lower than the HAZ making it the primary region for creep deformation

during PWHT to allow stress relief to occur.

Up to the 1950's, the highest ferritic alloy used in most boilers was the low

alloy 2~Cr-1 Mo. By the end of the 1950's through the 60's 70's and 80's alloy

development continued. The end result was P91. A 9%Cr steel with 1%

Molybdenum and various alloying elements. The steel is normally used in the

quenched and tempered condition and the starting microstructure is tempered

martensite.

7.1.1 General Chemical Composition 1

ELEMENT min% max%C 0.08 0.12Si 0.20 0.50Mn 0.30 0.60P 0 0.020S 0 0.010Cr 8.00 9.50Mo 0.85 1.05Ni 0 0.40

'V 0.18 0.25Nb 0.06 0.10AI 0 0.040N 0.030 0.070

7.1 .2 Mechanical Properties2


Tensile Strength 585-770 MPaElongation >= 20 %

Impact value (DVM) >= 80 JHardness Brinell 218 - 250 HB30

- -7.2 CHEMICAL ANALYSIS




Element New MaterialCarbon 0.22

Chromium 9.19Nickel 0.19

ManQanese 0.48Molybdenum 0.86

Vanadium 0.24Sulphur 0.001

Phosphorus 0.013Silicon 0.27

Titanium 0.01Copper 0.05Cobalt 0.005

Niobium 0.07Tin 0.01

Tungsten 0.005 ,

The chemical composition of the test material was within specified range for

P91 material. Only new material was available for testing.

7.3 SPIRAL NOTCH TEST RESULTS (Constant load, PWHT ThermalCycle)

Due to fact that no service-exposed material was available, a batch of the new

material was artificially aged to represent service-exposed material (described

in Chapter 3). The results from the spiral notch tests on the new and artificially

aged P91 are shown in Table 7.2 and Graphs 7.1 and 7.2.

Nominal Stress P91 P91

(MPa) New Material Artificially agedMaterial

407 Base Metal Base Metal356 Base Metal Base Metal305 Base Metal Base Metal254 Base Metal Base Metal

800

700

600

500

E~.3

400l!..Q.

E..>-300

200

100

0254

20

18

16

14

_ 12~c:0

'iii 1001c:0W

8

6

4

2

0254


Appendix 2.


seNice exposed material was studied by SEM to confirm the mechanism of

fracture. The fracture surfaces are shown in Figures 7.1 to 7.4.

Metallographic samples were prepared from new and service exposed

material after the CGHAZ simulation. The samples tested at 254 MPa (new

and artificially aged) were also taken and metallographic samples were



Condition State Figure DescriptionNo. \

7.5 Base Metal: Tempered martensite (quenched and tempered condition), with a dispersion of fineNew carbides in the martensite laths.

As- 7.6 CGHAZ: Tempered martensite with areas (colonies) of fresh untempered martensite.simulated Artificially 7.7 Base Metal: Tempered martensite (quenched and tempered condition). The grains are more

decomposed (tempered) than in the new condition.Aged 7.8 CGHAZ: Fresh martensite.

New 7.9 Base Metal: Tempered martensite.7.10 CGHAZ: Tempered martensite with some patches showinQ fresh martensite.

PWHT Artificially 7.11 Base Metal: Structure shows more tempering. The martensite laths are more broken apart andthere are fewer laths visible in the structure.Aged 7.12 CGHAZ: Partly tempered martensite.





1500

e. 450•1400

2350e1300

S 250'll5 200

-- As-Simulated: New

-- As-Simulated: Aged

-+-PWHT(50kg): New

-- PWHT 50k : A ed

The results from the spiral notch test were of a similar nature as the results

from the tests on X20 material. All the samples tested, both new and

artificially aged material failed in the base metal area of the sample at all the

different testing loads. The failure temperature of the new material for all the

samples tested, was higher than the comparative service-exposed material.

The amount of elongation per sample was higher for the new material than

the artificially aged material. The ductility of P91 at elevated temperature

compares well with the ductility of X20. The strength of the material is due to


carbides, molybdenum carbides as well as vanadium carbides.

From the SEM photographs it is evident that both the new and artificially aged

material failed due to a ductile fracture mechanism. Due to the notch a stress

concentration is present and the yield strength of the base metal is lower than

that of the HAZ during the PWHT. The base metal will be the primary region

for plastic deformation under the influence of a tensile stress until failure.

The structure of the base metal in the artificially aged condition is such that it

will be softer than the base metal of the new material. This leads to failure in,

the base metal areas because the yield strength is lower for the base metal

than the CGHAZ. CGHAZ thermal simulation resulted in similar structures of

partly tempered martensite with fine carbides within the martensite laths. This

is due to the very high hardenability of P91. A fUlly martensitic structure can

thus be obtained in these materials with slower cooling rates in comparison

with that of the two ferritic grades (%Cr-%Mo-~V and 2~Cr-1 Mo). During

cooling of materials with a high hardenability the martensite that formed first is

still at a relatively high temperature and longer periods of time and undergoes

auto-tempering as the work piece cools down to room temperature. The

structure of the sample that underwent PWHT (spiral notch test) showed a

more tempered state as expected due to the thermal cycle it was subjected to.

The hardness profiles showed that the mate!ial in the new and artificially aged

condition had almost the same hardness after CGHAZ thermal simulation

throughout the HAZ. This confirms the metallography that shows almost no

difference of the CGHAZ for both new and aged material. The artificially aged

material base metal showed a lower hardness than the new material as

expected due to the ageing process. After PWHT the hardness of the HAZ

was not significantly reduced, as was the case with X20. The hardness (yield

strength) is still higher than that of the base metal, which points out the fact

that the base metal is the area where deformation took place during PWHT.

P91 material creep strength is obtained due to the high content of chromium

as alloying element with molybdenum and vanadium in small quantities. The

hardenability of the steel is very high. PWHT restores toughness and ductility

and does not seem to increase the strength, as is the case with %Cr-%Mo-

'l4V. The yield strength of the base metal for new and service-exposed

material will always be lower than that of the HAZ making it the primary region

for creep deformation at elevated temperature to provide stress relieving.

1. StahlschlOssel - Key to Steel, 1998

2. Guntz, G., Julien, M., Kottman, G., Pellicani, F., Pouilly, A. and Vaillant, J.e.,'The T91

Book", Vallourec Industries - France, 1990

Private 8ag 401752022 ClevelandTelefax· (011) 629-5528Telephone (011) 629-5111

SANAS~

~5ANAS AC:REOITEOlABORATORYISOWoe 25/SAB5 0259 1'9901

and EN4500' "9891 Chemical Tec~~~gic.s Intemltionll

CERTIFICA TE OF ANAL YSIS

To: R LOOTS

MA TERIALS TECHNOLOGYSCIENTIFIC SERVICES DEPARTMENTTSI

Report No.:Logged on:Reported on:

WELDING A T CRAPAGED MATERIALS

These results are reported on an air-dried basis.

COMMENTS:

1& -pP Willie Delport .1U SENIOR OFFICER (COAL AND X-RA YJ

Phone no: 629-5254

This report relates only to the specific sample(s) tested as identified herein. The test results do not apply to any similaritem that has not been tested. .

Private 8ag 401752022 ClevelandTelefax (011) 629-5528Telephone (011) 629-5111

SANAS~

TSANAS ACCREDITED LABORATORYISO Gu"'. 25/SABS 02591' 99D}

and EN4SOO1 [1989)

Logged on:Reported on:

Sample 10: 200315390 200315392- -- ----_._----

1/2 1/2 1/4 NUUT 1/2 1/2 1/4 QUODescription:--------- ---- ._----- ----

-----------Units Value:Component: Value:

CARBON % 0.14 0.19CHROMIUM----- ----- % 0.46 0.53NICKEL % 0.08 0.16MANGANESE % 0.51 0.54-- --~--_.-MOLYBDENUM % 0.51 0.54VANADIUM----- % 0.28 0.31SU[PHUR"------ -- - ----

0.012% 0.01PHOSPHORUS % 0.008 0.014SILICON % 0.12 0.24TITANIUM % 0.01 0.01COPPER % 0.09 0.15COBAL"-- -------- % 0.01 0.02. - , .._ .. _-~ .._---

% 0.005NIOBIUM 0.005TIN----------- - % 0.01 0.02TUNGSTEN

--- _._-.----- - --% 0.0050.005----._------_._-- --_._-----

This report relates only to the specific samplers) tested as identified herein. The test results do not apply to any similaritem that has not been tested.


SANAS~

~SANAS ACCREOITEOLABORATORYISOGwcle 25/SABS 0259 1'9901

and EN45001 I' Ssg}


-------- -_._---_.Sample 10: 200315393 200315394Descriptfo-n:----------- 2 1/4 NUUT 2 1/4 OUD

..... ------.-- ---- -----------------------------..- .-. ------- --. __ ._- ---_._-------------------------------

Component: Units Value: Value:CARBON--- -_.._- -----.------ % 0.11 0.10CHROMllinn------- % 2.41 2.57Nic-KEI------- -- % 0.05 0.17MAN-GANESE % 0.49 0.54MOLYBDENUM % 0.87 0.91VANADIUM % 0.01 0.02SULPHUR % 0.001 0.003PHOSPHORUS % 0.018 0.021SILICON % 0.11 0.04TITANIUNi-------------- % 0.01 0.01COPPEFC------- - % 0.04 0.12COSALT % 0.01 0.03NIOSIUM----------- % 0.005 0.005;ffN"------- ------ % 0.01 0.01TU-NGSTE-N----------- % 0.005 0.02

This report relates only to the specific sample(s) tested as identified herein. The test results do not apply to any similaritem that has not been tested.


SANAS ACCREOlTED LABORATORYISO Guo<I.25/SABS D259 119901

and EfI,I45001 f19891


Sample 10:. ----- ----- -Description:

-_._---_._.- -----------------------------------------200315395 200315396

----X20 NUUT X20 OUD

Component:_. -.- .~- --_._-_ ... --_.".---

CARBONcJiROMiITNi-----NICKELMANGANESEMOLYBDENUMVANADIUMSU[PHUFf------_.- --- -,- - --PHOSPHORUS

_SILICON-----TITANliiM-- --- ---COPPER ---- --------COSALT---- ---NIOBIUMTINTUNGSTEN----

Units%%%%%%%%%%%%%%%

Value:0.22

12.050.750.590.820.26

0.0040~0160.070.010.110.010.010.01

0.005

Value:0.19

12.290.480.450.780.30

0.013 _0.0200.13M1---0.090.01

0.0050.01

0.005


Private Bag 401752022 ClevelandTelefax (011) 629-5528Telephone (011) 629-5111

SANAS ACCREOf1'EO lABORATORVISO G.>de 25/SAB5 0259 I' 9901

and EN45001 (1989)


Sample 10:- - ..- --- -_._._- ---

Description:200315397P91 NEW

~--------~~---------------Component: Units Value:

.... - _. _._~_._---_._.-._----CARBON % 0.22CHROM-IUiIt,-------------- % 9.19... _._----,--_ ..-. _.._---

NICKEL % 0.19_._- ---_. - ._-- ----, ..._-----MANGANESE % 0.48MOLYEfDENUM-- % 0.86VANADIUM % 0.24SULPHUR % 0.001PHOSPHORUS % 0.013SILlCO"'- .-------- --------- % 0.27TITANIUM .. ----------- ---~----------- -0-.0-1---

-- ._--- - --- ----------COPPER % 0.05C-OSALT----------- ---_. % 0.005NIOBIUM ---------------- % - 0.07TIN- -.--.------ - ------~-.---- % 0.01

- ---- - ------- -TUNGSTEN % 0.005



SANAS~

~s.e..NASAc:REO:TEQ LABORATORYIS: Guice 25/SABS 0259 (' 990)

and Er-v':S:D1 11989)


ALLOYING ELEMENTS IN STEELCARBON

ESKOM Method No 106 Rev 2 AccreditedESKOM Method No 119 Rev 1 Accredited


U 500~e::J~ 400CIlCo

ECIlI- 300

Appendix 2: Results from Spiral Notch Tests%Cr-%Mo-Y.•V New: 356 MPa

oo

4.5 m0"::JcoIII..•.

3.5 o·::J3'3

2.5 -

2

Time (h) .

-0.54

-Temperature-Elongation

Appendix 2 (Continued)%Cr-%Mo-%V New: 305 MPa

- 500o~~~~ 400CDQ.

ECD••.• 300

oo 2

Time (h)

4.5 m0-~ccIIIr+

3.5 0'~3'3

2.5 -

-0.54

I

j-TemperatureI-:ltJ~longation

o 500~f:Jni 400..8-ECDI- 300

Appendix 2 (Continued)%Cr-%Mo-Y4V New: 204 MPa

4.5 m0":J

CQIII

3.5 go:J3'3

2.5 -

oo

. -0.542

Time (h)

I

.-. Temp..erature.1••••••••Elongation .. J

o 500t...f::s~ 400GlCoEGlI- 300

Appendix 2 (Continued)%Cr-%Mo-y,N Service Exposed: 356 MPa

5m0'::s

CQDl..•

4 o'::s3'33-

oo 2

Time (h)

o4

l-....Tem.pe...ra.tu..reji___ Elongatioll_

- 500o~e~'tG 400..Q)c.EQ)

~ 300

oo

Appendix 2 (Continued)1!zCr-1!zMo-Y4VService Exposed: 305 MPa

5m0'~ccAI-4 o'~'333-

o42

Time (h)

i-Temperat. ureJ'I-Elongatiol1 _

IT 500~f::sf 400CIlQ.

ECIl•... 300

oo

Appendix 2 (Continued)%Cr-%Mo-%V Service Exposed: 204 MPa

4.5 m0'::scoIII..•.

3.5 o'::s'33~ 2.5-

2

Time (h)

-0.54

,-Temperature, I1-EI()1l9Citio_1l j

Appendix 2 (Continued)%Cr-%Mo-%V Service Exposed: 153 MPa

4.5 m0"~ccI»-3.5 o'~3"3

2.5 -

- 500o~f~1G 400~G)c.EG)~ 300

oo

-0.542

Time (h)

-Temperature i

-~Io,,!gation

- 500u~f:J'; 400•..CIlQ.

ECIlI- 300

Appendix 2 (Continued)2Y"Cr-1Mo New: 356 MPa

4.5 m0":J

(QIIIr+

3.5 o':J3'3

2.5 -

oo 2

Time (h)

-0.54

-Temp-erature i:- Elongation J

o 500~f::I-; 400..CIlc.ECIlt- 300

oo

Appendix 2 (Continued)2Y4Cr-1Mo New: 305 MPa

2

Time (h)

5m0"::IccIII

45-::I

'333-

o4

.-Temperature (OC. ) I

:-Elongation (mlTllJ

Appendix 2 (Continued)2Y4Cr-1Mo New: 204 MPa

4.5 m0";:]coDl..•

3.5 o·;:]

3"3

2.5 -

oo 2

Time (h)

-0.54

:-Temperatur.e. (OC.»[-Elongation (m_lTllJ

o 500~f::s~ 400CDc.ECDI- 300

Appendix 2 (Continued)2%Cr-1 Mo New: 153 MPa

oo

4.5 m0-::sccDl..•.

·3.5 o'::s-33

2.5 -

2.5

Time (h)

-0.55

-Temperature (0C)

-Elongation (rnm)

- 500o~~~E 400Q)Q.

EQ)

I- 300

oo

Appendix 2 (Continued)2Y4Cr-1Mo Service Exposed: 356 MPa

5m0'~

(QII)..•.

4 o'~3'33-

2

Time (h)

·0

4

,-Temperature:-Elongation

o 500L~::s'la 400..Q)D-EQ)

~ 300

oo


2

Time (h)

4.5 m0'::sccIII-3.5 o·::s3'3

2.5 -

. -0.5

4

.-Temperature:I

-Elongation j

o 500o-~:s1U 400L-ei)Q.

EeI)~ 300

Appendix 2 (Continued)2%Cr-1Mo Service Exposed: 204 MPa

4 m0":s

(QII)..•

3 o·:s3'32-


0-

o 2

Time (h)

-14

o 500t..e~'lG 400~II)Q.

EII)

~ 300

oo


2

Time (h)

4.5 m0"~ccIII

3.5 go~'33

2.5 -

-0.54

'-Temperature:-Elongation

[) 500~E::s1U 400•..QlCo

EQl

I- 300

oo

Appendix 2 (Continued)X20 New: 407 MPa

5m0'::slCIII..•.

4 o'::s3'33-

2

Time (h)

·04

-Temperature,

-Elongation

- 500o~~::I-; 400 -..II)c.EII)

.- 300


oo 2

Time (h)

o4

i-Temperature·[-Elongation

0- 500~f::::I

7U 400l-ll)Q.

ElI)••.• 300


oo 2

Time (h)

4.5 m0'::::IccDl••3.5 0'::::I

'33

2.5 -

-0.54

;-Temperature-Elongation

- 500o~2!~as 400•..G)Q.

EG)•.• 300

oo


5m0'~ccIII...•

4 o'~3'33-

2

Time (h)

o4

.-TemperatureI

i-Elongation j

IT 500!L.

f:J1! 400CDCoECD

to- 300·

Appendix 2 (Continued)X20 Service Exposed: 407 MPa

5m0":JccIII..

4 0":J

3'33-

oo 2

Time (h)

o4

I-Temperature i

I-Elongation

0- 500L.~~e 400CDc.ECD.- 300

oo


4.5 m0"~coIII..•.

3.5 o'~'33- ------~ - 2.5 -

-Temperatur. ej'-Elongation _.

2

Time (h)

- -0.5

4

0' 500~f~~ 400CDQ.

ECDI- 300


oo

5m0"jccAIr+

4 o'~3'33-

2

Time (h)

o4


[) 500t..~::I~ 400CDc.ECDI- 300


oo

o4

5m0"::IcoIII..•.

4 o'::I

'333-

2

Time (h)

-Temperature:I

-Elongation J

0' 500o-~:se 400CDQ.

ECDI- 300

aa

Appendix 2 (Continued)P91 New: 407 MPa

2

Time (h)

4.5 m0':sccIII..•.

3.5 o':s'33

2.5 -

-0.54

Il-.....Temp.eratur..e... 1___ Elongation J


6 500t...e~e 400 -CIlQ,

ECIlt- 300 .----.--- .....

oo 2

Time (h)

4.5 m0"~ccl»..

3.5 o'~3'3

2.5 -

-0.5

4

I-Temperature[-ElongCltic>n ..

[) 500t..2!:;,fti 400~CDQ.

ECD•.• 300

oo


2

Time (h)

4.5 m0":;,ccDl

3.5 go:;,'33

2.5 -

-0.54

I-Temperature[-Elongation

- 500(,)

~e::::I':G 400..II)c.EII)I- 300


oo 2

Time (h)

4.5 m0"::::I

lQIlJ-3.5 o'::::I

'33

2.5 -

-Temperature-EI0I!9~tion

-0.54

o 500~f!:2~ 400CI)CoECI)I- 300 --

oo

Appendix 2 (Continued)P91 Artificially Aged: 407 MPa

2

Time (h)

o4

-Temperature-Elorlgatic:m

o 500t..~;:,-~ 400Q)Q.Q)

E~ 300

o·o


2

Time (h)

5m0";:,

CQIIIr+

4 o·;:,

'333-

o4

.-Tempe ..ra.tureJ·- Elongati()'l_.

0' 500~e~e 400Q)Q.

EQ)

to- 300·


oo 2

Time (h)

5m0"~coIII0+

4 0'~'333-

o4

-Temperature!-Elongation

- 500o~e::J~ 400CDQ.

ECDI- 300


oo 2

Time (h)

5E'§.c

4 .2..C'lIC)Co

3 W

o4

I --~ 1

;-Temperature \:-Elongati()n __J


carbides as well as molybdenum carbides.

From the SEM photographs it is evident that both the new and service-

exposed material failed due to a ductile fracture mechanism. Due to the notch

a stress concentration is present and the yield strength of the base metal is

lower than the HAZ during the PWHT. The base metal will be the primary

region where creep deformation occurs under the influence of a tensile stress

up to the point of failure.

The structure of the base metal in the service-exposed condition is such that it,

will be softer than the base metal of the new material. This leads to the failure

in the base metal areas because the yield strength for the base metal is lower

than that of the CGHAZ. CGHAZ simulation resulted in similar structures of

partly tempered martensite. This is due to the very high hardenability of X20.

A fully martensitic structure can thus be obtained in these materials with

slower cooling rates in comparison with that of to the two ferritic grades (%Cr-

%Mo-~V and 2~Cr-1 Mo). During cooling of materials with a high

hardenability the martensite that formed first is still at a relatively high

temperature and longer periods of time and undergoes auto-tempering as the

work piece cools down to room temperature. The structure of the sample that

underwent PWHT (spiral notch test) showed a tempered structure due to the

thermal cycle as expected.

The hardness profiles showed the service-exposed material base metal had a

lower hardness than the new material. The hardness of the HAZ's showed

some difference. After PWHT the hardness of the HAZ is significantly reduced

as toughness and ductility are restored due to the tempering of the hardened

structure. The hardness (yield strength) is still higher than that of the base

metal, which points to the fact that the base metal is the area where creep

deformation was concentrated during PWHT.

X20 material creep strength is obtained due to the high content of chromium

as alloying element with molybdenum in small quantities. The hardenability of

the steel is very high. PWHT restores toughness and ductility and does not

seem to increase the strength, as is the case with %Cr-%Mo-'Y,N.Due to the

high hardenability, a X20 component during welding operations is held at a

certain temperature before PWHT. This could prevent full transformation of

the HAZ to martensite and retained austenite could be present before PWHT.

The temperature before PWHT must be such that it prevents cracking due to

the high hardenability and also to allow almost full transformation. The yield

strength of the base metal for new and service-exposed material will always

be lower than the HAZ making it the primary region for creep deformation

during PWHT to allow stress relief to occur.

Documents

SUSCEPTIBILITY OFSERVICE EXPOSED CREEP RESISTANT …